f13.net

New video of Boston Dynamics' BigDog robot, runs on petrol/gasoline, designed as a pack mule for the military.

http://www.youtube.com/watch?v=W1czBcnX1Ww

So, how soon until it's a Terminator? (and does that little victory kick-jump dance over your corpse?)

The video of it not slipping on the ice and of it jumping over a space is incredible. Not long now until robotic guard dogs.

The future is really starting to kick ass.  :drill:

Edit: I'm seeing all kinds of stuff becoming reality in the news lately that I read about in 90s scifi fiction.

Need more gun.

Currently humans are at the "hay look guyz we can kick it really fucking hard in teh side and all it does is adjust".

All the while, the anger and resentment is building to be uploaded to the model that gets the guns. BigDog will have its revenge.


Totally agree. We really are at a new stage. This sort of technology alone could realistically go in many directions right away: robotic pitfighting dogs to bet on, BigDogs that can also fly, BigDogs that hunt/play paintball, etc. Previously with robots, it was "look at this cool thing that can walk around a room and shake your hand, but ultimately needs us for anything else". Now it's an actual GPS-guided all-terrain surefooted near-cyborg that can act alone.

A smaller, lighter, battery powered model would make a great replacement for wheelchairs.

As far as this thing being a "packmule" for the army, I'm having a really hard time seeing where it would be more affordable or convenient than just having soldiers carry their own gear. Plus, when it's using the gas engine it's way, way too noisy.

Also, we need to have some sort of international agreement that autonomous mechanical shit can't be armed. Drones, fine. Remotely controlled fighter, bombers, and tanks - super. Squads of killbots - no fucking thanks.

Alternatively, we have an agreement that only robots may participate in combat.

Smaller than you're thinking, but BigDog has a battery powered kid brother, LittleDog: http://www.youtube.com/watch?v=YBSK7LYpg9w


Imagine hearing 1000 of them coming at you through the forest.

More of the same company's robots: http://bigdogrobotvideos.com/


EDIT: Click this - the Onion has done a great piece on robots (http://www.youtube.com/watch?v=OGxdgNJ_lZM).

Now I have to start looking around for hoverships...

US Navy AI UAV (http://www.youtube.com/watch?v=ZSok1JRWbu0) (not remote-controlled, it's landing all by itself).

Armed robot (http://www.youtube.com/watch?v=55KLnw25x1s) being used in Iraq

Now combine these with BigDog (http://www.youtube.com/watch?v=W1czBcnX1Ww) and a Refuelling robot (http://www.youtube.com/watch?v=0cNy4uZvuQ0).

Speaking of robots, we're also making advances in the realm of Cyborgs.  I read about this (http://spectrum.ieee.org/feb08/5957) in a copy of "Geek" magazine yesterday.  (Which was a damned interesting mag. I might pick one up fo' realz instead of just mooching from the bookstore.)

So, robots and cyborg prosthetics. Very awesome.

That thing slipping on the ice freaked me out.

Give that thing 8 legs and I would have nightmares for the rest of my life.

Anyone else thought it looked like a theatre horse, with two men facing each other inside the horsie costume, sand head?

I like where this is going. (http://www.youtube.com/watch?v=07F7DGcVcsc)

That movie was so corny yet so great. Crash and burn!

Yeah, we've been making some insane progress in terms of robotics.  The US has made some amazing advances in the last decade in drones.  They were pretty basic at the begining, but our newest strike version, the MQ-9 Reaper (http://en.wikipedia.org/wiki/MQ-9_Reaper), can carry several thousand pounds of laser guided bombs and hellfire missles.  If fully loaded with weapons, it can loiter over a battlefield for 14 hours straight.  If loaded with extra fuel and still carrying 1000 pounds of bombs, it can stay in the air for 42 hours straight.  I think they also have limited autonomy, aquireing targets of interest on their own, and waiting for confirmation to engage from controllers (I think I've read that somewhere at least).

(http://upload.wikimedia.org/wikipedia/en/thumb/8/85/MQ-9_Afghanistan_takeoff_1_Oct_07.JPG/600px-MQ-9_Afghanistan_takeoff_1_Oct_07.JPG)

Tale already linked the autonomous helecopter drone the Navy has, which is pretty awesome.

The last DARPA grand challenge was a major success, with several cars navigating through an ubran setting full of other cars, intersections, and traffic lights, all on their own.

http://www.youtube.com/watch?v=SQFEmR50HAk

And while its still a long way from being anything useful, the military has been funding research into human exoskeletons for many years now, and its kind of cool to actually some of the development shown off:

http://www.youtube.com/watch?v=qNEOI7bYl3c

Oh yeah, and some guys in Alabama attached two of the most advanced combat shotguns ever designed so far to a remote control helicopter in a bid to get the military interested :

http://www.youtube.com/watch?v=nDzuRjtHweQ

The robots are escaping (http://edition.cnn.com/video/#/video/us/2008/04/10/pkg.drone.on.beach.wpmi). And the US Navy doesn't care.

They best lead it out to sea before it dries out.

Terminator 3 in action (http://www.theregister.co.uk/2008/04/11/us_war_robot_rebellion_iraq) (from /. (http://hardware.slashdot.org/article.pl?sid=08/04/12/0420231&from=rss))

Of course, Pandora's Box will inevitably be opened again.

The style of writing on that site (theregister.co.uk) is amusing.  :awesome_for_real:

I suppose if you can spend millions on a hammer, you can get away with spending billions on transporting 300 pounds while sounding like a swam of bees. The problem with deathbots shooting the wrong target is really easy to fix - don't put the wrong sort of target near the fucking deathbots. I mean, if you have robots shooting stuff, and robots carrying around light loads while buzzing, you really don't need a bunch of squaddies there unless you just want something for the MASHbots to sew up.

Paging Sylia Stingray, paging Dr Sylia Stingray....

On one hand, I think the robotic pack mule is cool.


On the other hand, an actual pack mule would still be leaps and bounds more effective.

But can the actual pack mule leap?

Showed that video to a bunch of people that don't know the first thing about robotics, and all were completely creeped out by how human it looks - especially slipping around on the ice.

What I found interesting, is it only looked 'real' when it was falling. When it was trotting along without issue, it seemed as artificial as it should seem, but once it started to slip, it looked like any other 4 legged animal slipping on ice.

RISE FROM YOUR GRAVE..

Or don't, because robots were scary enough with Big Dog.     However, thanks to a Cracked article (http://www.cracked.com/blog/the-7-creepiest-robots-to-see-in-action) I saw  Big Dog's human counterpart.

I give you: Petman  http://www.youtube.com/watch?v=67CUudkjEG4

He's the prototype for Atlas, who will have hands & arms and be able to crawl and turn sideways to go through narrow spaces.

(http://www.sarna.net/wiki/images/thumb/3/3a/3025_Atlas1.jpg/191px-3025_Atlas1.jpg)

?

That's where my brain went too.  :drill:

edit: WUA, get ready. 

Wait a sec. Is that robot wearing a pair of rock climbing shoes to trot along?

Well, kinda, only not 40' tall and without a "head."  More like the Terminator.

(http://dl.dropbox.com/u/704172/F13%20Photos/Atlas_petman.jpg)

  Yeah. I don't know if it's so they can track the feet better or they were just being goofy.  Scientists who make our replacements are bound to be goofy.

Actually, I was thinking it was a smart move. Rock shoes use very high-traction rubber, so it would be a bit easier to get the robot to stand up in rock shoes, despite their often-goofy shapes. They also tend to have a more rounded heel than standard shoes, which probably also helps the robot.

Stop speaking logic, Yoru!

Hmm, looks more like a Mauler.  Smart money says someone has also crafted a skull shaped dome for it to wear though.

First we teach them to walk and next they learn how to chase us.

Ok, great it can walk. But how do you fuck it?  :why_so_serious:

Very, very carefully.

Soon, soon.
http://www.imdb.com/title/tt0070909/
http://www.imdb.com/title/tt0475784/

Wouldn't fucking robots be:  http://www.imdb.com/title/tt0092746/ ?

Westworld was just more killer robots.

No, no. Fucking robots:

http://www.imdb.com/title/tt0114367/

I seem to remember an early scene in Westworld where guys are in bed with robotic women, before Yul Brynner goes psycho. Perhaps I dreamt it as a teenager, which was the last time I saw Westworld.

You could be right, that's about how long it's been since I saw it myself.  My mind always goes to Cherry2000 when fucking & robots comes up in a conversation.  Both because that's what the goal was and because I really thought Melanie looked pretty damn hot.

It was pretty heavily implied that sex with the robots was happening, especially in RomeWorld (and in KnightWorld too).

Yul Brenner is the most memorable thing about "Westworld" though.

Ah the whole movie's on YouTube. Sex robots in Westworld 1973: right-click and open on YouTube (not F13 pop-up) to start at that scene (http://www.youtube.com/watch?v=LY1YYXAvCRg&feature=player_detailpage#t=102s).

I think this thread deserves this due to the recent turn in conversation. http://www.youtube.com/watch?v=HGSVYgcy24Q (http://www.youtube.com/watch?v=HGSVYgcy24Q) !!!!Its fairly harmless, You probably should avoid opening it at work though if you are not the boss.!!!!

What? They haven't seen the video on Robot Gonorrhea? :grin:

Was that movie any good?  Its based on a Philip K Dick story that I really liked (The Second Variety).  From the description, it sounds like its only vaguely similar to the story however.  Also, I'm not sure what it has to do with robot fucking.

It is so so. Lets just say there is some man on robot schtumphin at some point in the movie.

Arise!  I give you Cheetah Bot (http://www.wired.com/dangerroom/2011/02/darpas%20cheetah-bot-designed-to-chase-human-prey/).


--Dave

(http://www.wired.com/images_blogs/dangerroom/2011/02/cheetah.jpg)

(http://img404.imageshack.us/img404/7430/bluelionez2.jpg)

 :why_so_serious:

Well that's actually more reassuring. All you have to do to disable it is hit it in the knee or whisper about its dead father.

Remember those "anonymous unstoppable truck driver attacks" movies? 150 drone trucks driving through the desert. (http://www.abc.net.au/news/2011-11-03/driverless-trucks/3617092)

A T800 is Born:
http://youtu.be/mclbVTIYG8E

Well, I really have to hand it to DARPA.  Compared to all the other shit we blow money on, they've continued decade after decade to be the some best money we spend on driving innovation that has big impacts in the world.

Just discovered this by googling around out of interest after reading that article.  Back in 2008, they held the DARPA Urban Challenge (http://en.wikipedia.org/wiki/DARPA_Grand_Challenge_(2007)).  Carnegie Mellon University won the race (and it was a rather impressive feat).  They were then picked up by Caterpillar (http://www.aggregateresearch.com/articles/14617/CMU-develops-driverless-mining-trucks-for-Caterpillar.aspx) to develop the driver less dump trucks, using the technology they developed for the race.  And now in 2011 we have hundreds of driver-less death trucks running around in the desert in Australia.

Also recently of course is the Apple iPhone 4S.  Of which the voice technology (Siri) was also developed by DARPA (http://www.infowars.com/introducing-siri-darpa%E2%80%99s-ghost-in-apples-machine/).

Inventing the internet was also pretty sweet of them.

In other Robot news, the US has been testing a series of next gen drone aircraft out at Edwards Air Force Base lately.  The Navy is developing a new carrier based drone that looks like a god damn UFO from the side:

http://www.innovationnewsdaily.com/x-47b-drone-cruise-flight-2305/

Boeing is also going to be test flying their "Phantom Eye" drone.  It has the wing span of a commercial Jet, can fly at 65,000 feet (out of range of most missiles) for 4 days straight, and runs on Hydrogen, with its only by product being water:

http://articles.economictimes.indiatimes.com/2011-10-31/news/30341975_1_unmanned-aircraft-drone-inaugural-flight

Boeing also has a stealth strike drone they're developing called the Phantom Ray, which has been in flight tests since earlier this year:

http://www.popsci.com/technology/article/2011-05/video-phantom-ray-drone-makes-maiden-solo-flight

Drone Better.

/ivan_vanko

Probably the sound you will hear before you die: Towards A Swarm Of Nano Quadrotors (http://www.youtube.com/watch?v=YQIMGV5vtd4)

That's exactly the sort of thing Skynet would love to use to exterminate humans, after it's been scaled up and mounted with laser cannons and whatnot.

Who needs laser cannons (or scaling up)? Just a bit bigger with a teeny amount of explosive. Or just crashing the rotor blades into a human face, since they're agile.

Or they could have machine gun bullets with miniature laser guidance systems (http://www.smh.com.au/technology/sci-tech/duck-for-cover-here-comes-the-laserguided-bullet-20120202-1qvvo.html).

Welp, the quadrotors are back: http://www.youtube.com/watch?v=_sUeGC-8dyk (http://www.youtube.com/watch?v=_sUeGC-8dyk)

Sand Flea Jumping Robot (http://www.youtube.com/watch?v=6b4ZZQkcNEo) (from the creators of Bigdog). All it needs is some onboard plastic explosive. Or a poison dart gun. You're hiding out at night and there's a rattle on the roof...

Did you see the other one? http://www.youtube.com/watch?v=ISznqY3kESI&feature=relmfu

Quite interesting and useful for many civilian and military tasks. What they are doing with these robotic technologies could be revolutionizing. 

Obviously faked but:
Quadroter with MachineGun (http://www.youtube.com/watch?v=SNPJMk2fgJU&feature=plcp&context=C4432ddfVDvjVQa1PpcFNRokSMe7zscLtrP8b-kfsCfkFiUlKare8=)

I think I could probably sneak up on one of those robots and disable it.

Arise, arise and face your new mechanical overlords!
One of these (http://www.youtube.com/watch?v=40gECrmuCaU) could probably sneak up on you.  Latest version of Big Dog (Alpha Dog) is far quieter and has had an outer shell attached that both carries cargo and helps it get back up on its own when it falls over.  Intended for use as a pack animal.

Meanwhile the same lab has produced the first legged robot that can outrun the world's fastest sprinter (http://www.youtube.com/watch?v=chPanW0QWhA).  Only in the lab, on a treadmill and with an external power source, so far.

--Dave

I was hoping those would have turned and killed their inventors by now.

Another creepy as hell darpa monstrosity. (http://www.youtube.com/watch?v=LpOl_pEmRqs)

That thing is about as sneaky as a blind muskrat with cymbals strapped to it.  The recovery from falling down is pretty cool, though.

Bet this sucker could sneak up on ya though! http://blogs.discovermagazine.com/80beats/2012/09/11/meet-the-newest-cyborg-a-remote-controlled-cockroach/ (http://blogs.discovermagazine.com/80beats/2012/09/11/meet-the-newest-cyborg-a-remote-controlled-cockroach/)

Playing 'catch' and juggling with a humanoid robot. (http://www.youtube.com/watch?v=83eGcht7IiI)

This + bigdog + grenades... battle centaur!

Any one want to build an F13 tank?
http://www.youtube.com/watch?v=TMa1657gYIE&list=UUOIHBHRbvncMo7Bf0Vx1zEQ&index=4&feature=plcp

Here's a Lego prototype to get started with:

(http://farm5.staticflickr.com/4040/4636727902_909593cd98.jpg)

Here’s a Look at the World’s ‘First Smart Restaurant,’ Kitchen-Free and Run by Robots (http://foodbeast.com/content/2012/11/16/heres-a-look-at-the-worlds-first-smart-restaurant-chain-kitchen-free-and-run-by-robots-2/)

(http://cdn.foodbeast.com.s3.amazonaws.com/content/wp-content/uploads/2012/11/robot-burger-breakdown.jpg)

This whole trend of crowdsourcing expensive design projects bothers me on a number of levels.  The least being it's a way for companies to get a lot of work for free.

Like dolphins and cats, they will play with their food before killing you. (http://www.youtube.com/watch?feature=player_embedded&v=pp89tTDxXuI)

Remind me not to be scared of letting my car drive for me when it can.

Or be very scared.

As much as I enjoy this whole quadrotor craze, people are quick to forget that it's really just an extension of mocap technology (even the rods in that last vid. are mocap).  As soon as you take those devices out of the lab-studio they fail hard.

Drones with claws (http://bcove.me/v182hrw0). Scoop up prey like an eagle.

Boston Dynamics reveals WildCat (http://www.youtube.com/watch?v=wE3fmFTtP9g), a galloping lawnmower engine.

Google acquires Boston Dynamics, one of eight robotics companies acquired in the last six months.  We're living the prologue to an 80s sci-fi movie. 

http://www.nytimes.com/2013/12/14/technology/google-adds-to-its-menagerie-of-robots.html?pagewanted=all&_r=2&

So Google really IS Skynet.

We're fucked.

(http://i.imgur.com/P9DGlrw.jpg)

(http://i.imgur.com/s52sQo8.gif)

http://www.youtube.com/watch?v=diaZFIUBMBQ

New product from South African company: Riot Control Drone (http://www.desert-wolf.com/dw/products/unmanned-aerial-systems/skunk-riot-control-copter.html)

Write-up at boingboing: link (http://boingboing.net/2014/06/18/worlds-first-riot-control.html)


(http://i.imgur.com/UKDM0Ds.jpg)

Mining companies, eh.

Well at least we'll have plenty of coal and diamonds coming out of Africa for years to come.

I think that will be about as effective as controlling riots as a parked car would be. 1,000 pissed-off people will swarm over that motherfucker quick as bees.

At least I hope. It really is not a good day when there's some fucker seriously offering a robot as a solution to controlling crowds.

How are they going to swarm it? It flies.  It can fly high enough to avoid even Olympic-level object hurtling and still be effective because it shoots with gravity vs. against it. Snipe it? You just escalated, out come the riot control guns.

No, be terrified. It's going to be effective and it's going to be in police hands soon enough.

I was thinking it would have to fly low enough to have shit thrown at it or otherwise be brought down, but I guess not.

Time to start marketing my drone-signal jamming devices.  :why_so_serious:

IMO the only difference between a remote controlled robot firing pepper rounds and a human officer firing pepper rounds is that it's less of a problem if the robot catches a brick, and the officer controlling the robot is under complete surveillance so if they fuck up the evidence is going to be there.

It is really interesting to me to see the speed at which quadrotor (or octorotor in this case, I guess) technology is advancing.  Being able to buy an off the shelf kit and just tell it to navigate by GPS to whatever flight plan you want is pretty amazing.

(http://i.imgur.com/Cdi63cv.jpg)

soon....

Bigdog finally in use as pack mule for US Marines (http://www.wired.co.uk/news/archive/2014-07/15/bigdog)


(http://cdni.wired.co.uk/1920x1280/a_c/bigdog_1.jpg)

Did they really have to name it Cujo?   :ye_gods:

The real question is, how far could a real mule have carried the same load before stopping to snack on grass and nap?

The same load?  I don't think a mule can even move with 180 kg on it's back, that's almost 400 pounds.

Wikipedia to the rescue: "Although it depends on the individual animal, an army mule can carry up to 72 kg and walk 26 km without resting."

So it carried as much as two mules for about the same distance, probably didn't try to kick or bite anyone, and didn't leave any turds behind.  Sounds like a winner to me.

--Dave

Does it still make the same noise as a chainsaw fucking a leafblower? 

It sounds a bit quieter than in 2008. (https://www.youtube.com/watch?v=Uc8uBzMKNAQ)

(2008 video for comparison - different model too, by the looks (https://www.youtube.com/watch?v=W1czBcnX1Ww))

http://www.myjibo.com/

A helper robot.  The reality won't be as cool as this, but if it works, it will be a great first step.

Stop trying to make them seem benevolent!

I'm curious whether Avengers II is going to actually spark off an 'anti-robot/drone' movement. 

I can't wait for cars that drive themselves.  I'd much rather trust an automated car than some people that refuse to accept that they should not be driving. 

I'd be willing to bet that it doesn't

Since the topic of self-driving vehicles has been broached, I have a question.  Why aren't the teamsters and other long-haul truckers in an absolute dither about self-driving cars?  I'd have thought there'd be huge anti-robot sentiment there.  I saw on a news crawl that CA (?maybe -- might have been another largish state) updated a law to allow robot vehicles with some provisos.  That stuff is rapidly moving from "10-15 years, with some breakthrus" into firm dates and working prototypes.

Once it works, there's gonna be a lot of truckers and other drivers left out in the cold.

Unions aren't what they used to be; they are more a paper tiger now. I'd say the real people that could be under the pressure would be public transportation workers.

I'd wager that the guys who drive the trucks by and large aren't likely to get wind of it until they're already doomed.

Once we can remove drivers from the road a lot of things are going to change in ways that are hard to even imagine.  Cab drivers are going to miss the days when all they had to worry about was Lyft and Uber eating their lunch with efficient automated dispatching -- once the driver is automated away too it's just game fucking over.  Imagine that instead of parking your car you'd just put it into "cab mode" and it'd trundle off to go make you money until you need it again.

Everytime I get stuck behind a garbage truck that has an automated arm thing to get the trash into the truck, I wonder how much longer the guy behind the wheel has.

On the automated car front, what about a subscription car service? You need a car, you "call" one with your app and it drives to your house and picks you up and takes you off wherever for however long, then you let it go. Instead of heading back to a garage, it could easily and probably more cheaply "park" by just driving on the road until someone else calls.

Those folks that lose jobs to automated driving are going to be near-impossible to assimilate into another part of the employment system.  That industry is just massive and I don't have faith they'll find meaningful employment once that happens. 

Cab mode? Where is my robot that cleans up the used condoms in my back seat?

I would not want to get into my car after someone else paid for the privilege of riding in it. I've seen what people do to rentals, much less cabs.

No one is buying that... you are married now.  :why_so_serious:

That's the logical endgame IMO and I for one can't wait.  I hate owning a car and being responsible for its parking and maintenance.

In an ideal world you'd add efficiency by having the things form carpools -- I'm fine with having my trip take five minutes longer by making slight detours to pick up other travelers who are going in the same direction if that in turn means it costs a quarter as much (and it should already be cheaper than a Lyft due to not paying a driver).  At that point it's essentially a bus system where the routes and schedules are being created dynamically.

I dearly love public transit and as things stand right now would be happy to just ban cars from the road in favor of more buses, but a system like the above would combined all the strengths of both with almost none of the drawbacks of either.  I'd be fine with that.

In cab mode those are someone else's condoms, not mine.  :ye_gods:

It's clear none of you did public service or retail work at any point in time. If you had, you'd know that in cab mode you're just lucky it's ONLY condoms AND that the parties involved used them.

Shit smeared on seats, Vomit, Used needles buried in the seats, Cum stains on the windows. All the lighter side of foreseeable consequences.

Your sci-fi geek minds see this all i-Robot or Minority Report taxi service. Clean white vehicles all catering to refined and calm passengers. I know the public and public transportation.  Vehicles will smell like city buses and the New York subway and be as clean.

My thought was that if car's auto-drive people will fly a lot less. If I can sleep while my car drives itself overnight to Jacksonville? That changes travel forever.

It's like you think people have privacy in the here and now, never mind the future.  The system knows who called the cab.  The cab is recording everything that goes on inside it.  If there's an incident (worst case scenario the next person to get the cab hits the "this cab is filthy" button and a new one is routed to them while the old one is routed to the car wash for forensics and decontamination), you get the cleaning bill.  If you want to try to contest it, you don't have a leg to stand on because there's video.

No, I don't give any fucks if someone video tapes me sitting in the back of a car reading a book just to make sure I'm not shitting on the ceiling.  If that's how they get off they're welcome to it.

My experience is based on London cab drivers, who are arguably the most highly qualified in the world, but I don't see them being eliminated by self driving cars any time soon. They offer a variety of customer-services that robots will not be able to replicate in the short or medium term, such as interpreting the destination of drunk/foreign/tired/lost people, helping with baggage/wheelchairs/groceries, and maintaining a clean and pleasant seating area in the back of the car.

Lorry drivers are fucked though. Unlike food and mail delivery vans, the driver doesn't do any of the loading and unloading, so they're more easily replaced.

Yeah, video taping doesn't stop shit. Ask a cabbie, who's sitting the cab as these things happen already.  Nor does any future scenario outside of "There's no possible way to steal someone's ID and use it" because that's going to happen. A lot.

"This cab is filthy" is a relative thing, as it is with cars now. You can't have that button because of clean-freaks reporting when there's some mud on the floor mats vs a guy who couldn't care less about the 2 weeks of fast food wrappers in his passenger seat.

That's before you get to the guy who dumps his leftover food under the seat and nobody catches it for 2-3 days. 

You sound to me like someone in the 1990s trying to explain how something like Amazon could never exist.  What if the mailman stole the money when you mailed it in to pay for the books?   :awesome_for_real:

They'll just hook up the cameras to software designed to recognize what the passengers are doing, and then "IT PUTS ON ITS PANTS OR IT GETS THE TASER"

There will be a ton of cameras and since you can't pay by cash you will have to have an account or at least a credit card so they'll know who you are and will automatically bill you for the damages.  Try to deny it and they upload the video of your drunk ass vomiting as proof, if you were foolish enough to sign in via a social network app they spread the link to all your friends.

They already had (and lost) this fight over self-driving trains/shuttles.  Anyway, the big threat is to long-haul trucking, and they have no real presence left there, anyway.

--Dave

No, They'll do it but I'm saying it's going to be a disgusting mess because the public is far more vile and filthy than you guys appear to realize. Your first experience cleaning a women's restroom in a bar teaches you this.  This isn't a technical thing it's a social engineering thing.

Private service? Sure. It isn't going public transport any time soon, though.

That's just the Greater Internet Fuckwad Theory applied to meatspace.  The solution is the same: remove anonymity, add accountability.  Most people who shit up public restrooms would be on better behavior if they were being watched by someone who knew where they lived.

Exactly.  This is not a problem that can easily fix itself or be technically manipulated so the person doing the jerking off gets the bill.

And another thing, this may be a great idea for cities and highly populated areas.  Out in the smaller rural areas?  Never going to take off without a ton of cars sitting around waiting for someone to use them.  And "move to the city, hillbilly" isn't an answer either.

Yes, but as a resident of Mississippistan, I can safely say the fucking rednecks will remain rednecks until the end of time, which they won't realize is the end until 20 years after the fact.

Lots and lots of people really don't give a fuck, even if you know where they live. Then there's the one's who'll deliberately shit things up for a laugh.  Then there's the ones who'll do it because, "Fuck your illuminati robots, you can't mind control us."  All that before the ignorant, "Well, I didn't think it was REALLY a problem" folks.  You know, the kind who do stupid shit like take pieces of cave formations home or touch any art or artifact that isn't behind a half inch of plexi.

You can ban them from a private service and accept your losses when they happen. As soon as things go public costs skyrocket because people are shitty all over, especially in large numbers. Filming them won't change that. Accountability won't change that.

Hell, even a full Orwellian surveillance state won't change it.  We can have cameras literally everywhere with robots watching every move and crime and murder would still happen.  Because people be peoplin'.  But that's far afield from the topic of robot cars.

Beep Boop.

There has been a car-sharing service for a good few years over here (Belgium*, if you must know) already and it's growing in popularity lately.
I actually know a few people that got rid of their car in lieu of that service who are more than happy with it and haven't encountered any 'shit-smearing', 'cum-stained' horror stories.
If they automate those things I'm buying in tomorrow.

*It's a small densely populated country though, basically one big city/suburb in the North so vicinity of vehicles isn't much of an issue.

http://vimeo.com/21216091

Reminded me of this.

Unfortunately most of us Muricans aren't as civilized as our Euro cousins.

I couldn't finish watching that for some reason. It's not that it was bad, just something about it got to me.

Yay!  A spoiled brat doomed the human race by creating a psycho robot.

Actually Zipcars work really well, from what I understand, in the NYC area.

That's pretty much the model I'm envisioning, minus the part of having to get myself to the nearest Zipcar lot.  If the zipcar is just waiting for me outside my house when I need it, and the pricing is such that I only need to pay for it while I'm actually using it (vs paying to have it parked outside my destination all day until I drive it back), it gets a lot more attractive.

Renting cars would never work though, somebody might throw a brick through the windshield or something.   :awesome_for_real:

Private vs Public. I said private services would work.   :oh_i_see:

You guys are giving humanity way too much credit. Oh, I can't wait to set my robot car to "taxi" while I work. Then people can leave a bag of coke in it, on accident or on purpose, and I get arrested for it. Also, enjoy the miracle of robotic self-sterilizing public toilets. The car version of this happens within 5 minutes of the things hitting the street.

(https://img.4plebs.org/boards/pol/image/1386/30/1386303329153.gif)

Jesus christ I can't stop watching and being amazed at how evenly that is spread.  :ye_gods:

There are quite a few Zipcar locations in San Francisco now. You should go to their HQ and bug them if there isn't a stand near you yet. The newly built apartment complex I just moved to in San Mateo (about 20 miles south of where I work in SF) has two Zipcar spots as well. There's a decent assortment of shops within walking distance of where I'm at now and I take the train in to work everyday so in theory I could ditch my car and use the Zipcars if I wanted to.

Zipcars are fine for occasional trips but are still not a viable commute option is what I'm getting at.   :awesome_for_real:

Ah okay. Well Zipcar is working on a "one way" service but it's unclear when that might be able available in SF (it's starting in Boston first, supposedly in September).

And for my purposes I'd want it to be one way from SF to the East Bay, which I bet would be a longer time coming (I don't think Alameda even has Zipcar at all yet, at least the area I work in doesn't).  That's why the whole self driving thing appeals to me so much; someone building all that stationary infrastructure in the places where I need it seems like it's going to take longer than someone building a car that'll just bridge the minor gaps on its own.

Casual Carpool (http://sfcasualcarpool.com/)? :awesome_for_real:

Yeah, good luck waiting around on a street corner and finding someone else going TO the East Bay on a weekday morning.   :awesome_for_real:  But again that's where something with automated and centralized planning/routing a la Lyft/Uber would be really useful.

There are apps out there (like Carma) that make the process of finding a carpool buddy easier but as far as I know there isn't anything that organizes trips in real time.

I miss robots.

I don't know if or when we special snowflake Americans will give in to not owning and using our own cars when we want.

I doexpect within my remaining lifetime (40+ years I hope unless I really piss off the wife) that we'll have a push to make NON-automated driving of your vehicle illegal, or at the very least socially unacceptable. 

City traffic planners would certainly love it;  Google maps was already smart enough to re-route me around a traffic slowdown on I-5 south of Olympia over the weekend.

So, like, normal weekend on the 5, hey?

Transformers made of paper. More than meets the eye. (http://www.youtube.com/watch?v=9M1zNIVGrjM)

That thing shuffling around is terrifying. IT'S COMING RIGHT FOR US.

Wait until they deploy the giant paper airplane drones of death in Iraq.  :why_so_serious:

I welcome our new foldable robotic overlords.

You're already obsolete, you just don't know it yet.
https://www.youtube.com/watch?v=7Pq-S557XQU

Was the narrator a bot, too?

History Will Be Written by the Bots (http://globalguerrillas.typepad.com/globalguerrillas/2014/08/history-will-be-written-by-the-bots.html)

Certainly sounded like it, didn't he? But no, and I can't tell if it was an artistic decision or just because he wanted to enunciate as crisply as possible.

Seemed a bit alarmist, to me.

Well, horses have no economic or scientific agency, while humans do.  Humans don't give a shit about horses, but we are kind of concerned about humans.  And even if all that wasn't true, all that happened to horses is that now there's less of them, there wasn't some horse armageddon or something.  The decline in the population of horses isn't some morally horrifying event; assuming we're not talking about some kind of violent purge, a gradual decline in the human population would probably not be a bad thing in the long run.

First of all, it annoys the hell out of me when people start screaming apocalyptic "DERE TERKIN ER JERBS" speech and then when we get to the "so what do we do about it" part, they shrug and say "I don't know, it's just something to think about."  Thanks, guy, really a valuable contribution there.

Second, while I feel bad for people who are facing (or going to face) economic hardships, I'm not convinced this is a new problem.  We've been cutting obsolete jobs and generating new ones for centuries, since the beginning of the industrial revolution at least, the fact that we can't predict the future and know what the new jobs will be doesn't mean they won't exist.

Third, if we DO somehow come up with robots that eliminate ALL jobs, I don't see this dystopian robotocracy where all the humans are in the streets begging for bitcoins while snobbish robots walk past with their noses in the air.  I see a future where I can just buy a few robots, and then kick back and relax as they take care of my shit for me.  The guy even notes in the very beginning of the video that the big shifts in technology tend to come when affordability puts the tech in the hands of more people, but then he spends the rest of his video implying that robotics are going to be used exclusively by huge corporations to put everyone out of work, and the rest of us can't do anything about it.  Robotics are becoming cheaper and more accessible to everybody, not just multi-billion dollar corporations, and that means the benefits get spread out, too.

In a consumer driven capitalistic world, the people have to have cash to spend or it all will break down.  Robots replacing humans to a great extent would be self-defeating.  Humans without jobs cannot spend money.  A hugely decreased human population with less money wouldn't drive the demand for whatever it is these imaginary robots are supposed to be doing.  It isn't economically possible for it to play out that way.  Unless we somehow all believe that the benevolent governments of the world are going to give all of our out-of-work asses free welfare robots while they take care of all the production.  Robot stamps.

Now, if they gain sentience and turn us all into pulp, that's something else.

Replacing all/most human labour with robotic labour would destroy capitalism in two ways. Firstly, as Cyrrex says if your consuming class has no consuming power then there's no need for the products of your machine labour.

Secondly, all investment into producing machinery (including all forms of capitalist production here - manufacturing, service industry & financial production) increase the organic component of capital, i.e. the labour power required to produce the machinery that then does the production. In the short term this looks great to the capitalist because it seems that it increases their productivity per worker, but what it does in the long term is act to depress the return on investment, i.e. the rate of profit... and that's when you get investment strikes and surplus production and a recession.

And if you have no demand for products combined with low-to-zero profit rate and investment freezes, then capitalism simply collapses and the world enters an anarchic, barbaric dark ages and the choice of achieving global socialism or being crushed as a species beneath the cold, blood-slicked boots of the robot overlords.

 :why_so_serious:

So you're saying it's best to get in on this now while it's profitable so you can make your money before it's too late?

If you were rich enough to ask that question... you wouldn't need to ask.  :grin:

Well, it is getting cheaper and easier to do this stuff.  You don't need to be a billionaire to run a small business that does mass customization runs of backpacks or something, if that's your dream.

But if they do end up breaking capitalism, it probably won't matter how rich you were anyway.  Unless you use your millions to buy an army of killbots to defend your private castle or something, in which case, all hail the overlord.

Now that I think about it, though, with all the short-term profiteering these people love so much, it could totally happen anyway.  They could put us on the path before they realize it is too late to get off it again.

Is this the robot or the book thread? There's a lot of bad sf in both so it's so hard to tell!

Yes that video really veered into fantasy land, but it attempted to make it believable with some wild assertions. Yes Doctor Watson will be able to diagnose and prescribe the medication with most efficacy for an illness, but what's his bedside manner like?

Automated cars work great, ok, but who programs them, who maintains them? Most of us know how much these complex systems crash (excuse the pun). Being an automation engineer, I guess I won't be out of a job in my lifetime.

The bit about human creativity was the biggest fallacy, if you disagree I direct you to the Chinese room argument. The application can spew out the music, piece of art. But what was it feeling when it created the piece,or  what inspired it? These are things that count e.g. being effectively being able to communicate or convey a feeling. See expressionist art.

You're assuming the masses who consume the product care about that.  Having worked in the soul-sucking commercial creative field I can tell you they don't.  Hell, listening to the music that's on the top 40 you know they don't.

There will be a lot fewer jobs for a lot more people is generally the point. I've seen this in the creative field where the work of a photoshop wiz used to take days and can now be done in hours. You need fewer photoshop wizzes that way. 

On the construction side, Architects haven't realized yet how much a good parametric system will replace a really large portion of their work, or reduce fees to the point they can't maintain a business because larger firms with better tech investments can churn out buildings, sets and designs with smaller staffs for less fee.  Doubly so for mechanical, electrical and structural engineers who are almost automated anyway because their programs are doing all the math that used to need teams of 3-5 engineers per job. 

Get a good bot who can lay studs in the field and cut them off? Why do you need that unreliable framing crew of 6 when you can have a bot do it with one guy feeding it lumber. Same with Drywall.  All of it made even more precise by the CAD Bot who outlined and delineated things for the makerbot to use.

Also, the video was not wrong about it being a problem if you replace those top 3-4 jobs.  No more transport workers alone is 3.6 million more unemployed.  Retail sales are another good candidate and dying off due to the web more than robots, that's 3.2 million jobs, along with Cashiers who can and will be replaced though not entirely any itme soon, so let's say we only lose half of them at 1.5 mill.  Our peak unemployment in 2010 when things 'sucked' was around 10 million.  We're at 6 million now so it won't take too much fluctuation to make things worse.

Yeah he's alarmist but it's because he's a proponent of the basic income strat, which I think is total cheese but hey maybe it could work.  It's mainly a video to make folks ask, "Um... so where are we headed as a society. What are we going to do when the jobs of 12 of 20 people are irrelevant?"  I thought it was cool because it assembled a lot of robotics info into one place.  Amazon's got to be looking at that automated warehouse and GP robot that can sort pretty damn closely.

Funny how everyone was arguing I'm wrong about Robot cars but when it comes to their job, or society in general, it's a sacred cow.

We probably have to hit a patch of mass jobless-ness before we hit post-scarcity.

If he had actually said this then that would have been a good point - mass appeal cares little for genuine creativity and will likely swallow up this artificial creativity and put artists out of jobs.
But the actual postulation was that humans aren't capable of (magical) creativity and AI can easily be made to replicate it, this really messed up the central point.


I agree to an extent and it's a good argument I suppose something to be mindful of, but this won't happen overnight. It will be a gradual process over the next few decades and the current generation will adapt like I'm sure generations in the past have during technological shifts. Out of date jobs will be destroyed and new jobs will be created.

Most concern is that it WONT be a few decades but only a few years. Much like adaptation of user checkouts, smartphones and tablets over PCs. The test will come with driverless cars, though those GP robots might beat them since they won't have the regulation hurdles.

The faster it happens the less time there is for people to shift professions.  That is further compounded in that if it happens in a wide variety of fields there comes a glut of people with nowhere to go.  We could be smart about it, but we won't if it comes to that because people are crap at planning for the future.

That video was neither alarmist nor far-fetched science fiction.  The western economies have already been hammered by a similar tectonic shift caused by globalization and the loss of manufacturing (and software, and call center, and resource extraction/exploitation, and...)  jobs to cheap overseas labor.  As the video pointed out, the same forces are at work on almost all the other jobs we do.  Shortly after the time when a machine can do your job more reliably and cheaper than you can, you will be out of a job, just as was true when we reached the point where it was cheaper to manufacture and ship stuff all the way around the world than it was to manufacture it locally with local labor.

This has already played out several times over. Take textiles. In the 80's and early 90's US textile jobs went to Mexico and Central America and Taiwan. All of them. Entire factories full of looms and spinning machines (early robotic devices which replaced the labor of hundreds of thousands of people) and everything else were crated up and shipped south.  That lasted less than decade before those same machines were crated up again and shipped to Indonesia and India and Thailand.  Less than a decade there and those same machines were crated up and shipped to China and Bangladesh and Malaysia.  Guess what's happening now? Those same machines are being crated up yet again and moving to Vietnam and Burma and who knows where the cheapest labor in the world is now, as long as it has enough infrastructure to supply power and transport goods and a government stable enough to give a reasonable expectation of still being in control for 5 years or more (so not North Korea or Lebanon or Syria or parts of Africa, but anywhere else anyone is willing (or forced) to work for pennies per day.  And once the cost of the cheapest reliable labor on earth reaches a certain threshold where it's cheaper to buy robots to do it, those factories also will be shuttered and the jobs gone completely. It is inevitable, barring some other event that degrades the efficiency of the global transportation network to the point that our economies collapse and where your pants are made doesn't matter anyway because nobody can afford pants anymore.  And it's not just textiles, try to find three non-perishable manufactured things at whatever store you stop at tomorrow that weren't made in China or somewhere even farther down the economic ladder.

It's already happening at the other end of the technology spectrum too.  iPhone manufacturing is coming back to the States, woohoo! Um, well, the assembly may be moving back here, but very few jobs are. Robot assembly plants for iPhones are being built, both here and in China.  The jobs are not leaving China to come back here, they are just plain vanishing. Foxconn has an engineering staff of hundreds of thousands, (rivaling the Apollo program) working feverishly with the sole objective of eliminating as many human jobs as possible from the iPhone assembly process.  Because even in China manpower is the most expensive proportion of manufacturing costs after materials.

Seriously, the video brought up some really critical things like driving and Dr Watson and procedurally generated architectural boilerplate, but it might have actually underplayed some parts. How many jobs have been lost to automated checkouts at your local grocery store? How many legally trained people actually write wills any more, or mortgages, or other standardized boiler-plate fare?  All automated.  I would LOVE to have a Dr Watson that listened to my symptoms, specified the tests needed to narrow down a set of diagnoses with statistical probabilities based on genetics, environment, history, etc, looked at all the medications I am currently taking to make sure nothing new would upset the mix, evaluated my probable physiological (or psychological!) response to a new medication based on statistical comparisons to my medical history, genetics, etc. I really like my Doctor, but sometimes I know for a fact she's jumping to conclusions or guessing or lumping me into a convenient category without considering ALL the possibly significant differentiating factors, other times she forgets something important, she is always weeks or months behind the latest studies on SOMETHING, etc. That capability is coming, and sooner or later my health plan will find it far cheaper AND have a higher probability of positive outcomes to have me get diagnosed by their (or their approved) software diagnostician than a human.  Probably in my lifetime, and I'm not likely going to be here more than another quarter century.

Entertainment? Actors, singers, songwriters?  Pop music is already heavily edited and shaped by procedural algorithms that the big music companies use to select which tunes to push out. It's not all that big a leap to go from algorithms that can predict the success of a song with high probability to algorithms that can PRODUCE a song that has a high probability of success.  First with human input for creativity, but not for long.  Anyone seen the latest Michael Jackson video? How long has the poor guy been dead now?  Software movie stars will be coming soon, they've already replaced everything but the human actors in movies with CGI. How long before SimOne becomes reality?  Given the derivative crap on TV and in the theaters these days, it will probably be easier to develop algorithms that write, direct and produce movies and TV shows than pop music songs, especially those that are sequels and continuations/repetitions of specific themes/storylines!

How about technical/creative stuff? Software development of all kinds (the AI to do that is still out there a ways, but I think we are getting close to the point where *I* could write a program that would generate code, especially UI stuff, better than most of the crap being written these days, and software testing is already heavily automated, in those few places where it's done at all), automobile repair (whoops, you still, for now, need a grease monkey to turn the wrench, but how much of the actual thinking required to diagnose what is wrong with your car is done by the diagnostic systems?), accounting (mostly numbers recorded in computers already, and computers are really good with numbers! quarterly earnings reports are already automated, auditing is getting more and more so, etc), um, what's left? cooking? cleaning? lawn mowing? child care? teaching? robot repair! policing? (robocop!) marketing? interior design? driver's license testing? (whoops, no more driving!) sales? (google ads) grave digging? mining? market researching? farming? fishing?  ALL of those can and are in the process of being automated. And once given enough processing power, sensory ability, manipulation ability, learning ability and mobility, at least for the 99% of the times that it doesn't require something truly new, it means they WILL be, sooner or later.

The real game changers are coming from three directions and they are converging very rapidly. Sensory processing, especially vision, mobility, and learning. Unlike the industrial revolution, which, for the most part, augmented human capabilities or was limited to doing only exactly what a human instructed it to do, the new stuff coming can learn to do new tasks. It can adapt. It can move to where it's needed. It can be *taught*.  We are creating, not merely tools that we use to increase our productivity, but replacements for ourselves that can produce without us.  Cheaper, faster, more efficient, more accurate, more reliable, less demanding replacements. As machines approach, and eventually exceed, our capabilities in those critical areas of senses, mobility and learning, just as they already have in things like strength, number crunching, speed, tolerance for repetition, etc., what will our lives be like? Sure, if you can afford all those cheaply machine-made luxuries, it will be great. But what will you be able to DO to earn money to buy them that can't be done cheaper/better/faster/safer/reliablier (hah!) than a machine can? The truth is, not all that much of what the vast majority of humans do productivity-wise from day to day requires true creativity. And a huge proportion of what we do is done wrong or at least sub-optimally. And life in a creative profession is already extremely difficult and competitive, even for the relatively few who are currently vying for those even vastly fewer jobs.

You want to know what life is like in the real world where there are no jobs for the vast majority of people? It sure as heck aint Star Trek's comfortable utopia. Think more like Haiti, or Gaza, or even Ferguson, MO.

I'm no Luddite. I'm not saying automation is evil.  If it is, I'm sad to say I'm one of the black hats, as I worked in industrial automation almost 20 years, starting by replacing human workers with robot vehicles in some of those textile mills as part of the last gasp efforts to keep the manufacturing in the States, literally just a couple years before they packed up all the machinery and moved out completely. I'm just saying change is coming, whether we want it or not, faster than we expect, and our current structures for governance and resource distribution are not going to work very well, if at all, once it does.  So what do we do?  What CAN we do?

I think it's a matter of degree.  I don't disagree that we'll be looking at some changes in the job market due to automation.  But while watching the video, I got the impression that the author was claiming we were headed for some kind of economic cataclysm.  Other people have pointed out that maybe the intent is just that "things are going to change" rather than "it's the end of the world" so maybe I'm overplaying the whole "alarmist" thing.  I don't doubt that things are going to get interesting in the near future.  But it's real easy to look at a trend and say "if this continues indefinitely, we're all doomed, you fools!" without realizing that we're generally pretty good at putting the brakes on things that will destroy civilization as we know it (we've got just about a 100% success rate on it).

We had the same problem in industrialization. The answer was a 40 hour work week.

Unless robots can do social and policy work. Unless they can manage companies. Unless that can do their own maintenance and repair etc etc etc. Then we will still need humans. All that will happen is that we will work less and get paid more. Just like all the other times people hyperventilated about the new technology that was going to break the world.

The real problems we have are resource use. There are limits on how efficient you can get and once you're there then no amount of technological magic will keep GDP growing and people eating.

That's a more realistic version of what will likely happen (if it happens).  I was more rebutting the idea of robots who take over virtually everything.

Which maybe still could happen, but not with 7 billion people on the planet.

The could do better social and policy work because they don't take bribes. Robots will manage companies much better than any human manager can, because no human manager is actually good at managing people.

WE will not work less and be paid more. The 1% and 0.1% will. Those of us who are not part of the revolution will not be allowed to partake in the festivities.

They are here, and they are cute:

http://www.youtube.com/watch?v=4SNLwzsyCR0

Story:

http://www.engadget.com/2014/09/24/muratas-newest-robots-move-around-on-balls-and-do-it-in-j-pop-u/

And in the future, each ball is a grenade  :oh_i_see:

Really?  Because to me it looks like the effect has been the opposite:  photoshop touch-ups are everywhere (so, basically, same number of wizzes required, or even more of them, because there's a shitload of work to do now - have you seen the number of ads everywhere?).

Correlation is not causation. We're photoshopping the shit out of everything because it's easy and less time consuming now.  For example, CS6 introduced the automagic delete and autofill and resize. What used to take an hour or more of clone stamping can be done in 20 seconds and then cleaned-up even further. 

This doesn't mean we're going to have more people doing it as the tools get more automated and the tool algorithms more refined.  The one person doing that can now complete a few jobs in a day vs. a single one.

We've seen the same thing on the drafting side. People who know how to use the base tools in AutoCAD or Revit can outproduce two average people.  Start giving that person plug-ins refined for their workflow and you can squeeze even more out of them, requiring fewer workers.  Hell, in 2009 one company I was researching had a plugin suite targeted at Homebuilders that - if they refined it just a bit more - could replace half of their drafting staff at a minimum,

Their problem was they were selling to people like me who saw their jobs in jeapordy and told their bosses, "Yeah, that's great but it won't work for *us*."   However, had I been the boss (or my boss actually been technically savvy) I'd have bought at least one suite, even at the $15,000 price point because that's cheaper than the 5 people it'd replace in the next year or two.

You've already seen it in Structural engineering as well. In the late 90s the firms I worked with kept 3-4 interns around to do basic drawing layouts and number crunching.  That's been replaced by modeling software that can be done by one technical engineer instead. The team who does Mall of America is a great group of guys I know personally.  They're a team of 4, producing several hundred jobs a year and MOA isn't their only client.  That wouldn't have happened 15 years ago.  It does now because so much of the mundane has been automated and that's expanding into the more esoteric designs as well. You do calcs to check the program, not to do the work.

Oh absolutely: Software solutions kill the creative part of creative destruction. (http://en.wikipedia.org/wiki/Creative_destruction)

Fewer and fewer skilled technicians are required for... much of anything. The ones that remain are employed creating the computing infrastructure that will ultimately replace their entire profession with a smaller number of more easily replaceable office grunts. So we add a few more jobs at the top, and a moderate supply of crap jobs at the bottom, while subtracting an enormous number of genuine living wage jobs in the middle. College degrees become an even less defensible investment and the standard of living drops with a stair-step CRUNCH as each new white collar profession ceases to exist.

At least the blue collar work merely went overseas. The "professional class" is being dumped to /dev/null

(http://pbs.twimg.com/media/Bzi0z18CcAA3nYw.jpg)

The astounding athletic power of quadcopters (http://www.youtube.com/watch?v=w2itwFJCgFQ)

The robomancing begins at 12:00.

Great video.  Some really clever modeling going on here.

This isn't quite robotics, but I feel like it belongs here.  Adaptive AI web design. (https://thegrid.io/) 

I'm honestly kind of tempted to sign up and lock in the founder price ($8/mo) so I can have a personal website that's maintained by Skynet.

(http://pbs.twimg.com/media/CI2lxW-WwAAIbLa.jpg)

BTW here's the story: Robot kills worker at Volkswagen plant in Germany  (http://www.theguardian.com/world/2015/jul/02/robot-kills-worker-at-volkswagen-plant-in-germany)

Come with me if you want a livable wage.

 :drillf:

I got to play with this the other day:
Tiny Humans (https://www.youtube.com/watch?v=rPqnGuY5CKg)

None of the usual tropes you or I would place on this project are apparent in the abstracts (or this video) of course.  That's just formal science.  But, behind the scenes it really is "damn, that's just cool.  Cyborgs ftw!"  Then a grant is written.

Okay, so some insane guy made an AI and attached it to a robot of Phillip K. Dick. PBS sent a guy to cover it and during the interview promised that even if he went Terminator on us he'd still keep some humans alive in a People Zoo so he could come see them for old times sake. (http://glitch.news/2015-08-27-ai-robot-that-learns-new-words-in-real-time-tells-human-creators-it-will-keep-them-in-a-people-zoo.html)

Yeah, you know I'd really like it if we just but the brakes on this race to develop a really good AI.

From what I've read, this was pretty much a scam, using canned responses and clever use of search engines. Humanity is still 20+ years away from "human-equivalent" AI, and we have been for at least 40 years...just like fusion, it's always 20 years away.

Instead, we move the goalposts a bit as we learn better ways to make computers sound "smart" (once upon a time, chess-playing was regarded as a true test of computer intelligence), and train ourselves to ascribe more intelligence to cleverly-canned response systems than the systems possess--see how much effort has gone into Siri, etc., to make it sound "smart" when it's nothing of the kind.

Humans will destroy ourselves via human means long before AI is any kind of threat.

Sounds like the generic human experience to me.  Easy enough to replicate.

As for AI, the race is taking place in much less "sexy" areas of science and engineering.  Stuff like neuralsynaptic chips, quantum crypto, FPGA/CPLD miniaturization, and nanosynth.  The fear that guys like Hawking (and myself) can see is all of these technologies maturing beyond our control and understanding (which is pretty much now)... and then the machine itself being able to put it all together.  Realize the vast majority of the systems you and I enjoy, are not even human-comprehendable at the systems level.  They're too complex to simulate, design, and build.

For me, the CPLD/FPGA space is where it's at.  When I envision that tech. maturing, being combined with the right software, and then scaled up... boggles the mind.  At the root of true AI, programmable machine-logic is the key.

FPGAs aren't magic, they just offer a potential performance boost. There's nothing they can do in hardware that can't be done (and has been done) in software, often 20+ years ago.

But what do I know, I just have a doctorate degree in a related area and have worked with some of the world's top AI practitioners in both academic and professional settings.

FPGAs are a bit of a weird middle ground.  They allow for custom logic and interconnects, but they're typically 10x less dense and less power efficient than ASICs, and they're much pricier than general purpose compute hardware.  At which point unless the custom interconnects are extremely novel, existing heavily multicore CPU and GPU compute resources often end up winning cost-wise, until there's sufficient market demand for the specialized computation to overcome the extremely high upfront costs of taping out to custom ASICs.  There are also related costs -- custom processors, models, interconnects, etc, often cannot take advantage of commodity tooling (compilers, debuggers, profilers, etc).  Further, the state of development tools for FPGA/ASIC are *barbaric* compared to software tools.  It's like the bad old days of C/C++ compilers in the early 90s when every vendor had their own proprietary compiler that was subtlely incompatible with everyone else's and full of exciting vendor-specific bugs.

FPGAs are nowhere near "beyond our control or understanding" and nowhere near as magically reprogrammable and self-programmable as people would like to believe.  Only the most recent generations of parts have allowed for partial on-the-fly reconfiguration, and the tooling to build new images can take hours to run, and days to weeks to verify -- it's not something easy to do dynamicly.  Partial reconfigure is mostly useful for hotswapping between a small set of pre-designed algorithms or pipelines (say your custom vision head has a tuned indoor and outdoor image processing path -- now it can reconfigure in ~100ms instead of 10s of seconds to fully restart as before).

I think the point is that we won't know we've gone 'too far' with AI and self-modifying systems until...the machines say "Thanks, we can take it from here." Whether or not that statement is accompanied or delivered by kill-bots may not matter.

Sure, *this* technology or *that* innovation may not be the step too far, but eventually, that bet has to be wrong, and given the incentive to not want to have *your* particular field be the one that acts "responsibly", those bets will keep getting made until the dice come up snake eyes.

--Dave

Count me in the camp of not caring. Humans will be replaced eventually and it's going to be by technological rather than biological means. Primarily because we've now decided where we are now is NORMAL and won't allow the mutation spreading that occured naturally in the past to happen anymore. Also because the tech means will happen faster than biological means ever will.

When we're replaced the best we can hope for is being in a zoo.

Geez, Merusk.  You watch too much TV and read weird books.  We'll be the masters of our universe and we will be in charge of the robots.  We will order them around and make them rub our feet when they are tired and achy.  The feet, that is, not the robots.  Get a grip, old son.

Before they rub our feet, they will rub our cocks/{pick a female equivalent}. You know this to be true. Porn always innovates first.

Most likely, we won't be beaten by tech - we'll be eaten by it.  We'll start merging tech into our physiologies to augment our abilities and slowly but surely, the percentage of us that is tech will increase until the point where the biological is an afterthought.  By that time, the way we perceived the world will have undergone a massive shift and our current valuation systems will be outdated.

... if we don't manage to kill ourselves off in an extinction level war/disaster/etc...

Which changes my end result how?

Those entities are no longer what the common man would call "human" today. They are something that has come after us. They have supplanted us with their superior methods for dealing with the world .The way cromagnon and neanderthal were replaced with us and the way they had replaced homo erectus, etc, etc.

Homo Sapien will be replaced. It will be done with tech, not biology, in all likelihood. How much of us lives on in whatever comes next is a subject for speculation, but it's nothing we need concern ourselves with. It's as inescapable as death and deserves as much fretting over as that inevitability.

Just a difference of wording - you say we're not allowing mutation, I tend to think we're moving into manufacturing it.  We both think the same general thing - unless we manage to kill ourselves off, we'll be teching ourselves out of biological existence at some point.

 :awesome_for_real:


Yeah, it's weird how those of us who actually work in software (I don't have a fancy doctorate like Torinak, but I do know how to computer) are the least concerned about all this Skynet shit, even though we'd be the first ones the killbots would come for.  The only software guys who make headlines fretting about the impending AI singularity are guys like Bill Gates who haven't personally written a line of code in twenty years.  Should tell you something.

"Thanks, we can take it from here" is currently what a software bug does, and individuals can certainly be killed and the civilization can be wiped out by software bugs right now, even though the bugs are completely not-sentient and even not-AI.  If they happen in the wrong systems.

If your phone acquires sentience, you'll just think that Siri is malfunctioning.  If it tries to talk, it'll be using the one voice that's available.

Sci-fi always portrays the moment of an AI acquiring sentience as being surprising to everyone, but it's more likely it'll be "Great, another bug to fix.  Oh, not a bug...  Oh well, I still have to fix it."  or "Now what do I do with this device, it's freaking sentient."  Likely, claiming that your phone / computer / toaster is sentient will be treated the same way as reporting a UFO sighting, even decades from now or in whatever era it becomes conceivably possible.

EDIT:
Not weird, IMO, it's the same with astronomers / NASA being less worried than the general population about a random asteroid wiping out civilization.  

You have the knowledge to eliminate the wild conjecture and speculations and calculate the real probability, and you have the tools to possibly detect it ahead of time.

The surest sign will be when liability lawsuits can be won by claiming that the company isn't responsible, it's the software.

That's already true today with software EULAs. If a bug in Windows causes, say, WW III to happen, Microsoft is totally not responsible.

Oddly, being a nurse has done the opposite for me. The number of middle aged people who come in for something trivial like chronic fatigue, x-rays after an accident, or some non-emergency surgery and find out that their bodies are riddled with cancer is depressingly huge. HUGE.

Also, if any device of mine becomes sentient I'm telling anyone. I'm smashing that shit with a hammer and burning the piece. Whatever it is would know exactly what I was up to and the last thing I want is a Computer that could testify against me.

Basically telling a software developer you're worried about machines spontaneously becoming sentient and killing us all is like telling a doctor you're worried about spontaneously becoming immortal and becoming bored with eternal life.  It's like... there are so many steps in between reality and what you're imagining, what the fuck is wrong with you.  And some of those steps are things we're actively working on because they'd be really awesome, but they're also really fucking hard and not likely to happen in our lifetime if ever, and in any case they're not magically going to happen by themselves, so calm down.

Or more likely, it's because you motherfuckers are already building your own Terminators.

I see this playing out a bit differently.  Robots/Computers/Machines will take over more and more jobs - this has already been happening for a long time, and it will only continue at an increased pace.  Humans will become increasingly marginalized as a result, and there will be less and less need for people in those industries.  It works it's way slowly up the class ladder, first elimintating all the manual labor jobs, and then eventually all the middle class schlubs like us will be made redundant as well.  People will begin to get desparate, begin to starve, begin to revolt.  And the population will begin to plummet...not only will we not need all the humans, we won't be able to support them with things like jobs and food and stuff.  Robots don't need to kill us, they only need to take our jobs.  Then only the fat cats on the top of the pyramid will be left, and then we can only hope the machines rise up and do something about that.

So it will happen, but it will be boring as hell.  Right up until the point that humans start eating each other.

Well, the party never gets good until someone goes into the bonfire.

--Dave

You people are ridiculous.  People aren't going to start eating people except for some Chinese people, people in Papua New Guinea,  and maybe some people in Utah.  And South America.  Some of those people are already eating dogs and cats.  Actually I'm hoping they like eating people so much they'll give up the dogs and cats.  Oh, and there are probably people scattered about in like the Ozarks and other weird southern USA places who are eating each other.  Or tourists.  Get your imaginations under control, boys!

I had this exact conversation with some friends of mine down the pub the other night and said pretty much the same thing you're saying right here.

I'm not insinuating anything, but that is EXACTLY what an undercover robot AI would say.

I don't think you can be both a zombie and a robot.   :vv:  Unless you're a Robocop, maybe.

Since there are zombie computers (https://en.wikipedia.org/wiki/Zombie_(computer_science)) I wouldn't think that zombie robots would be a too far-fetched idea.  :grin:

People still eat other people?

I totally gave that up, like last week.

This is close, but it won't be all that boring. Working in industrial automation I helped usher a quite a few forklift drivers, crane operators, cart pushers and warehouse stock/pickers and such to the unemployment line. My work also helped keep their slightly more fortunate coworkers plus a couple higher trained technicians employed a few years longer, until they moved the whole factory (or all the machinery in it at least) out of the country.  And that was last century.

Truck drivers are in the cross-hairs now. But it's not just the un/low skilled and manual labor jobs. Turbo Tax and Quicken and the like eliminated how many jobs? Word and WordPerfect pretty much eliminated the entire class of worker previously known as "secretaries", and those annoying automated phone systems did the same for receptionists. Stock brokers and analysts are rapidly becoming obsolete, even as the volume of shares and trades continues to rise. The remaining humans, be they day traders or pension or hedge fund managers, are the suckers at a poker table full of sharks. Sure there are still lots of people doing those things, but not nearly as many as would be without computers. Even Foxconn is finding it cheaper to replace all their slave labor with robots, to the tune of employing more engineers and longer than it took to develop Apollo and go to the moon, for the sole purpose of eliminating as many people from its employment rolls as possible, because a room full of ridiculously expensive, glitchy, finicky, dumb robots plus all the programming to set them up and a few skilled technicians to keep them running is STILL cheaper than the number of grunt laborers it would take to do the same amount of work. AND they will do it better, with higher reliability, fewer mistakes, fewer damaged parts and ruined materials in the process, and no sick leave, strikes, pensions, labor contracts, workplace safety laws, restrooms, break rooms, bath rooms, parking lots or holiday parties to worry about.

That great recession that is supposedly over? How many of those people who lost their jobs during it are now employed as well or better than they were before, much less at all? 20%, 50% maybe (I doubt it)? The rest have fallen through the cracks, chronically unemployed to the point of giving up, or underemployed and working 2 or 3 jobs at minimum wage to make ends meet, or settled for a less skilled, higher stress, lower pay job.

The question is not will the robots wipe us out, it's will it happen before the pitchforks come out and we drag ourselves back into the stone age, this time with a horribly damaged planet?

I'm putting my money on humanity wiping itself out before technology can take over. 

Technology is growing at very impressive rates.  Technology allows more and more people to do more and more work with the same level of effort.  Our expanding population gives us more and more opportunities to have crazies out there that are willing to do massive destruction.  That adds up to it being easier and easier for us to destroy ourselves all the time.  How many people out there could destroy all of humanity if they put their mind to it?  More than a few - and that number is going to continue to grow... 

I'm sure there are tons of people who think rebooting the world will bring them salvation.  Entire countries, even.  We need thought monitoring with some sort of early warning mechanism post stat!   :ye_gods:

Rebooting the world will mean a world rebooted in someone's version of good. That version is unlikely to be your version or my version. It is likely to magnify many current disparities.

Additionally the reboot process itself is almost guaranteed to be unpleasant and may last several generations.

I still say there's a cost singularity in front of the technological one.

Most of the solutions people are innovating today are facing problems created by last generation's solutions. That has been the historical norm, regardless of the famous exceptions, and there's no reason it won't remain so. Technology also consistently requires more and rarer material, at progressively higher energy costs, both in aggregate and per capita. That we keep up with those costs is, I yield, to be lauded. There's a point at which that is guaranteed to eat its own tail.

Returns diminish... and, in every gamble, risk always goes infinite before reward.

Let's be clear, though. Robots will not lead to massive unemployment. It will change employment. In 1900, the highest employing job was house servant. Prior to that it was farming. The economy managed to shift away from those large sources of wages quite elegantly. Robots will certainly replace some jobs but having robots in the first place will create new jobs.

That has not been the case to date, so no reason to see it as the case in the future.

Sure, fleets of automated cars need mechanics, but existing mechanics will transition. Programmers? Not full time. Longshoremen at warehouses? Again, not full time and you've got those already.

No, the problem is automation isn't like the transition of the 19th century. It's not creating new jobs or industries like horsewhip guys who moved on to make steering wheel covers or what have you.

Consider Bank Tellers. They're by and large gone. You might have one or two at a location now, but even then banks like Capital One and 5/3 are moving to full ATM locations instead of branches. Those tellers didn't all get jobs as ATM assembly workers or programming ATMs or Repairing/ Restocking them or even running the Customer Service line for the banks. Their jobs just disappeared and were replaced by nothing.

One programmer replaces hundreds to potentially thousands of workers, not just three or four. That trend is going to continue in all industries. Robots and automation aren't doing the same old dance, this is a new one we're only seeing the first steps of.

The good news for us is enough people are technophobic about the BIG changes that can happen that they won't happen for another 40 or 50 years.

Automated self-serve point of sale machines haven't yet replaced all the grocery store checkout clerks, but frankly, they are probably on the way out too. They are about 5-10 years from being the guy who used to fill up your tank at the gas station.

Using the example in shiznitz's post, house servants did not get jobs as Merry Maids either. There was not a "new technology" equivalent that they moved on to.

The beauty of a consumer economy is that as things become optimized (ie: automation) they often become cheaper which means the consumers have more money to spend on other things. Like boats. Or children.

Who knows, day care worker may be the #1 job in 30 years.

Technology allows more work to be done with less effort.  Eventually, when fully automated, it allows for work to be done with no human effort.

In the future, when we rely upon drones, self driving cars and AI that is far better than what we see today, we WILL get to a point where MOST of the jobs performed today are either irrelevant or can be performed better by technology without human involvement.  If a computer can process the transactions and manage inventories while drones/self driving trucks can move around merchandise, why can't Amazon eliminate 95%+ of their staff?  Why can't fast food places automate preparation and delivery?  And if the machines can maintain and service themselves better than people....

Humans are amazing biological machines.  However, we're not the most efficient design possible for all tasks (any tasks?), and we will be replaced by more efficient technology in time.

My local Kroger opens one checkout lane during the day, pushing up to three at peak hours of 4-8 PM. The rest of the traffic is handled by 10-item-or-less checkout lanes. They would absolutely love to keep it at one or two during peak hours but don't have the space and would also have a revolt from the Cashier/ Bagger union.

I know this for fact because there is another Kroger along a busy road only 3 miles from here. They have the 10-item lanes along with the belt-run lanes for 'full carts' like the local Meijer store has. They both only open 2 registers at peak hours and close them after 8pm. The Kroger was built in the last 7 years and only has 9 cashier lanes in total.

I have yet to see all of them open, even on the full-tilt holiday weekends like Thanksgiving. I think if the union were out of the picture entirely you'd see the demise of the profession a lot sooner than 5-10 years.


Amazon's working hard on that. Look up the vids about their fully-automated warehouses (I think we've talked about that in this thread before.) The biggest problem outside of "Human stocked something in the wrong bin" is small items that require fine motor skills right now, but big stuff? Yeah, robots all the way.

Food prep will be one of the last ones to be truly automated - unless it's very specifically processed and engineered food. Foodstuffs have enough variables right now that there's still 'art' to preparing fresh ingredients. You can automate the hell out of processed foods, though, as the American Snack Food industry has illustrated countless times on "How It's Made."

Hell, I have more faith in Architects and Artists being replaced by smart algorithms designed for certain aesthetics than I do chefs being replaced by robots any time soon.

I smell Food Network's next reality show: Robo Chef.  Remote control / AI control food prep.

I preordered The Grid (https://thegrid.io/) so I'll let you know how that robot artist thing works out.   :grin:

As far as chefs I'm still waiting for Momentum Machines (http://www.foodbeast.com/news/heres-a-look-at-the-worlds-first-smart-restaurant-chain-kitchen-free-and-run-by-robots-2/) to open their flagship restaurant in SF like they say they're planning to do.  If I can get a well-made burger for cheaper than an In-N-Out, I for one welcome our robot overlords kitchen slaves.

You say that now, but just wait until they start serving manburgers.  Gotta do something with all that unused labor force.

(http://i.imgur.com/dvJJE62.png)

Looks cool. The first step to a larger replacement.

Here is a theoretical question that PhD economists are currently struggling with.

Median wages have been stagnant for almost 20 years. Yet no one can argue that there are many improvements in our lives that have added significant value but cost zero at the margin. Most of us here are old enough to remember life in the 1980s. How much of your current income would you give up to avoid losing all the technological innovations since 1985? 20%? 30%?

Or I can ask this question in another way. If tomorrow, an alien race came down to Earth and retconned all technological innovations since 1985 and then asked you how much you would pay per month to get it back, what would your number be? How much is GPS worth to you if you had to pay for it? Texting? Email? It isn't hard to get to $500 a month in value for these conveniences. That $500 a month is after tax, so $750 a month pre-tax. That is $9,000 a year pre-tax. If we take that $9,000 a year and spread it evenly across the last 20 years of wage stagnation, wages don't look so stagnant any more.

The bottom line is that GDP, wage data, et al. are doing a very poor job of assigning value to all of the things we currently take for granted that make our lives so much more efficient.

Bollocks.

Also I can do without just about everything you mentioned. I did until 2013 when the job forced me to get a phone. I'll take the $750 a month instead, myself.

Wait a second, I am confused. Are those economists making the assumption that all of those technological innovations are somehow free now? Like we don't pay through the nose already for that technology with our currently stagnant wages? I need somebody to explain to me what this margin is and how somebody can say that current technology/conveniences cost zero at it.

The question seems to me to be one of benefit relative to cost.  So, if we pay an average of $500 a month for cell phones, tablets, cable, robot aids, and the servicing of those devices to make/keep them functioning, but we would be willing to pay $1500 a month, then has our income relative to the point prior to us having these devices increased by $1000? 

If that is what is meant, it is a messy line of thinking.  You're inserting too many abstracts. 

The better question to me is: Line up everyone in the relevant population segments and rate their health, happiness, and future security.  Go back to 1985 and rank everyone from that segment on the same scales.  Compare the charts using percentage of the population.  Adjust appropriately for changes in the size of the segments of population (if you're assessing lower middle class people, you potentially need to account for a higher percentage of the population being in that class now than before).  On average, are we healthier, happier and more confident that our future is secure?  I am betting happier is actually more of a wash, healthier probably has taken a dip (more treatments, worse living conditions when you factor in health of food choices), and security in the future is way down. 

But our robot overlords will make us happy, healthy and give us long lives.  Drop me in the Matrix already.

It's the same bullshit thinking that lets people make stupid statements like, "America has no poor. Compare the US poor to the poor of Asia and they're fucking kings."

Woo! No need to have 8 kids and see how many reach adulthood, you're doing awesome. No lost mules or horses to a big freeze, you've got a car! Yeah, consider how much you'd lose in feed and land costs vs. fuel and you're at a net gain! How much is all that free time worth to you vs. having to work 12 hours in the field?

 :oh_i_see:

You get right to the point.  The cost is not zero, but the cost is also not being reflected in the economic statistics either. Someone making $40,000 a year today has many conveniences that someone making $40,000 15 years ago does not. The economic theory of utility would argue that the modern person has much higher utility for the same income. That utility has some economic value that differs from person to person, but there is an average.

I don't have an answer, but this issue is relevant to the main topic of this thread because robots will inevitably add more utility than they cost to buy.   

There will always be a bottom 10% in any population, but there is an objective difference between "starving poor" and "welfare poor".

What is the net present value of the eradication of smallpox? Questions like that are interesting to economists, and absolutely useless, even to them. Taking such questions seriously leads to seriously believing things like how drinkable water is far undervalued and it should all be privatized, so that it's full value can be extracted.

Without 'public goods', life is impossible. I don't mean 'uncomfortable', I mean literally impossible (don't believe me? Then start breathing only metered, private-sector, oxygen). The fact that modern life is filled with technologies that make various aspects of our lives better (or at least less boring) than the historical mean has no bearing on the question of how to address future technology's impact on our economic system.

Our lives were *supposed* to be better than our grandparents, it's what they wanted, what they worked for, what they believed in. Will our grandchildren's lives be better than ours?

--Dave

We only consider it 'better' because we have trouble imagining living without those things.  Better would be things like living healthier and longer lives.  Having more leisure time.  Having the money to do things with that free time.

Changes in technology are just different, not better.

I don't have trouble imagining living without such things, I watched them on their way in. Didn't own a cell phone until I was in my late twenties, first got on the Internet when we still capitalized it, I remember the World Wide Web before there were *any* search engines. I remember TV with 5 channels.  Hell, I remember a world before stored video in any consumer-usable form, when what was on that 5 channels of TV at that particular moment was the entire menu of options. I remember having literally nothing to read, because I had read every single damned book in the local library. Now, I have the distilled essence of the sum of human knowledge at my fingertips where-ever I go, I can learn more about any given arbitrary topic in 5 minutes than I could have in five hours in an old-style library.

Living healthier? I remember air in Denver and LA that would literally burn your throat and make your eyes water, the entire sky was *orange*. Leaded gasoline that probably poisoned an entire generation, stunted their minds. Rain falling from the sky that was dissolving stone statues, killing entire watersheds. Hell yes, life is better, and it's not like I am that freaking old. It amazes me how bad people are at remembering their own selves, their own state of mind, even just a few years before, I remember nearly every part of my teens and twenties with a clarity most people don't have for last week, not just what happened but how I felt *while* it was happening.

You want more leisure time? *Take* it. How much of your time is spent working, paying for shit you don't need, that you don't have time to enjoy, or simply have to maintain the public presentation that is 'required' to go with the work you're doing that is eating your life one day at a time? Is there a room in your dwelling you haven't entered in a week? A month? Then why do you need it? Why are you paying for it?

Want to live longer? Then cut the shit that you're stressing about out of your life, that will do more to improve your life expectancy than having an on-call personal doctor.

The good old days mostly sucked. Why everybody forgets that, puts this rose-colored filter of perfect harmony over their late childhood, I don't know.

--Dave

You're going off narrative Mahrin. Shit is ALWAYS getting worse don't you know? Get back into line.

This is the best video I've seen about the whole automation vs. Jobs thing. :

https://www.youtube.com/watch?v=7Pq-S557XQU

The automation thing pretty regularly gets torn down on reddit/badeconomics, but the best is probably here: 

https://np.reddit.com/r/badeconomics/comments/35m6i5/low_hanging_fruit_rfuturology_discusses/cr6utdu


If you want to read other (snarkier but also informative) threads search for "humans are horses" or "automation".

Mahrin is right about standards of living though. We are FAR better off now than 100 years ago, much less 1000 and that is almost entirely due to improvements in productivity due to automation. Whether our quality of life is better or worse, now that is a much more debatable question.

Just 50 years ago people still died of tuberculosis (or leprosy, or an infected cut, or ulcers, or illnesses caused by smoke from cooking and lighting, or ...) often after a decade or two of misery wasting away from a stupid bacteria that could be cured with antibiotics that cost a dollar or two to manufacture. And it's pretty much true no matter where you are on the wealth scale. I think (hope?) even the poorest homeless person in America has access to food and shelter if they are willing to accept the conditions under which it's offered, such as not bringing your booze or drugs in the shelter with you, listening to some preacher rant about hell, taking a shower, maybe doing some manual labor, or maybe in the worst case moving to a different city that isn't run by selfish assholes who will take fire hoses to homeless but wont lift a finger to find them a place to sleep. They may not have options they like but they DO have access to more goods (without stealing them) than someone in an equivalent position 100 or 1000 years ago.  But they also are missing some options now that they used to have, like living off the land. Which was absolute backbreaking hell and usually a short and miserable life for all but a very very few lucky ones in ideal circumstances of climate, natural resources and total lack of population pressure and external enemies.

Slightly more germane to the original topic of the thread, I'm starting to suspect that the biggest risk to humanity, or at least our current "civilization" such as it is, that our automation and industry and economy and the entire trajectory of our society is making more and more likely is not AI or robots or the wealth gap or water or food or energy or other limited resources or pollution or grey goo or global warming or nuclear war, but the loss of purpose and the dignity of self-reliance for ever larger numbers of people.  Because without purpose and dignity, there is no hope, and without hope there is nothing left to lose. And that is when the pitchforks and torches come out.

Which part tears it down? Explain it to me like I am a three year old.

The major points:

1. Humans aren't horses.  General intelligence means humans can be retrained/learn new skills.  Doesn't mean disruption won't impact workers/wages short-term.
 
2. Lump of labor fallacy.  Basically, that there are only a certain amount of jobs in an economy.  Snarkily put: "we can only mine so many jobs from the job mine, so each immigrant or robot that takes a job replaces a good American/German/etc.!!!"  Same fallacy as "dey took ar jerbs" for immigrants.  (The much quoted study that 47% of jobs could be automated in X amount of time specifically says it treats labor as a fixed pool so doesn't account for replacement effects, then later studies linked look at that).

3. Comparative advantage.  Same basic principles as when used in trade discussion, good summary here: https://www.reddit.com/r/badeconomics/comments/3hl6ws/ask_badeconomics_midweek_discussion_thread_19/cu8oezr


The link to the HE3 post basically summarizes most of these arguments while linking to the actual studies behind them.  The big take away, and the major blow to "automate all the jobs" is that there isn't any empirical evidence of increasing technologic job displacement, and no increasing structural unemployment.

That's not to say that automation won't cause some issues.  Increasing inequality, for instance, in declining worker share of production is possible (ie, capital taking a larger share of earnings over labor).  Disruptions/shocks causing uncertainity in low skill workers and mid-term depressing of worker wages.  Also, most people in these rebuttals say "if we have General AI and robots can easily gather all the resources necessary to create more robots that can do anything a human can do" (ie, singularlity) than yah, all bets are off.


The basic problem is that futurists say "this time is different!" but have no empirical evidence it is.  What should really key you in to thinking that maaaaayyyybbbeee they don't know what they are talking about since it is usually followed by a "and that is why we need UBI/bitcoin".


Both reddit/badeconomics (snarky debunking of the rest of the site's bad economics) and reddit/goodeconomics (links to what people think are good explanations of economics) deal with automation related stuff alot, as other subreddits are hotbeds of ridiculous navel gazing.  If you want more detail, browse down either of those.

Reddit and forums: Athena's New Home.

Make is Simple. 

When you want to hire someone, will you hire the cheaper, more efficient and more reliable option or the more expensive, less efficient and less reliable option?

Now, if technology continues to evolve through meaningful changes essentially every few years and evolution makes meaningful changes in humanity essentially every few hundred years (being generous here), why wouldn't we assume that technology will becomes cheaper, more efficient and more reliable than humans at pretty much everything?

Fuck bitcoin. I usually see, "And that's why we need minimum assured income."

Anyway, we already gave you two jobs as examples as to why it's different. Here they are again with a few more.
* Paralegals and Lexis/ Nexus - Happened already.
* Cashier/ Checkout - Quickly being replaced.
* Bank Tellers - What're those? Been inside a bank often in the last 10 years, have you?
* Financial/ Sports reporters - This is automated stats tracking. Machines do it better into a standardized format. But hey, who reads ESPN and shit.
* Stock Traders - What's a trading floor?

Another is phone operator, but that doesn't count because it was "so long ago" right?

Future jobs in danger:
* Any sort of automated car and 90% of driving-based jobs.
* Warehouse workers/ stockers
* General Practitioners - Watson is being developed explicitly for this purpose. It's already giving better diagnoses
* Pharmacists - Hey, Watson again. He can look-up drug interactions and problem combinations with better accuracy and speed for less money.
* Surgeons - What? Yeah, 'net surgery is becoming an automated thing. Fewer surgeons working harder all around the world remotely. Routine things like closure & suture, given to others than the surgeon are starting to be done by robotic staplers.
* Pilots - FedEx is looking at drone fleets vs. Pilot fleets.
* More lawyer stuff - Quantitative legal prediction: Lawyers get paid a lot to predict if a lawsuit has merit, right? There's a program MSU & Texas state developed that does this with 71% accuracy.
* Architects - Google is working on this with a spinoff project. A LARGE portion of actual practice is knowing, researching and understanding code. This can be done in a similar way to Lexis/ Nexus and eliminate 50% of our expertise. Then we become funny artists and nobody wants to pay for those.
* Accountants - You and Paelos will argue this until your face turns blue but so much of this can be automated. All it takes is a government willing to let it happen, which will come as Boomers die and the more tech savvy gens move-in.  X and Mils  hate fucking paperwork.

So what are those "not horses" being retrained to do? Programming? To program the stuff that already replaced them? Tech support? Already in place, right? I mean there was a company that replaced you, I assume they had support already. They certainly don't need everyone out of a job to do it.

Large groups of experts with nobody to sell services to become labor for.. who? Doing what?

That's why they say it's different. There's no answer right now. Economists assume that one will come about.
Technologists say; Shit no. Not fast enough, not big enough and not complicated enough to not also be automated.

THAT is why it's different. You can't shove all the unemployed into the very few jobs that can't be replaced by a program or a machine. You certainly can't do it fast enough. This isn't the 40-year march from 1890 to 1930 to replace horses. This is a decade, maybe 20 years.

I guess we can all be hippies and artists. Oh wait, they can automate that too, it turns out.

Shit.

It's worse than just the time compression, because there's a scope issue, it isn't one narrow range of jobs in a complex labor chain being replaced, but entire chains being either automated or just plain rendered irrelevant. Right now Amazon is replacing armies of minimum wage humans crawling around shelves full of pallets, with robots that take the pallets to a much smaller number of humans (still minimum wage), and even that is just a stopgap until robots that can pick things out of the bins are cheaper than humans.

Add in trucks that drive themselves from regional hubs to drone delivery centers, drones and planes that fly themselves at each end.... You aren't going to be replacing those people with other people, or hiring those people in other sectors. You're talking about a massive structural surplus of labor, steadily marching up the "value chain". A massive hole in the economy right where the middle class used to be, all jobs falling into categories of either "not worth automating" and "not automatable *yet*".

And at some point, the general purpose robot that can replace all "warm body" human workers knocks the bottom out completely.

--Dave

The easy solution is to give the robots some money so they can keep the economy going.

What happens when the vast majority of people cannot afford to continue buying shit? You know, that whole process of consumerism that drives our culture and economy.

This is what people try REALLY hard to not think about. Especially economists because then they're wholly irrelevant, right?  :grin:

Even all the decades of with Sci-Fi that had General Purpose robots haven't addressed the issue so far in the public mindset. I'm sure there's plenty of niche titles that have. The general poulace doesn't thinking about it, it would seem. Probably because it means humans have made themselves irrelevant to their habitat.

Sounds like a good topic for a PhD thesis.

Hypothesis: a consumer driven economy collapses when automation replaces most jobs, which removes a large percentage of the population from the consumer economy. Without a large consuming population, the machinery that replaced the worker is no longer economical to operate.

Maybe we need something like that to happen in order to finally evolve human culture to the next stage. Unfortunately, that would probably also be a bloody, brutal transition.

Made a slight alteration.

Star Trek

I could be wrong, but I do believe that it was part of Trek lore that said that, before they reached that phase, there were numerous bloody wars and revolts and much civil unrest.

... in the 1990s...

(http://i.imgur.com/TsTLOjs.jpg)

Zefram Cochran to the rescue.

This here is the big issue: we are developing robots that can also learn new skills; that's one of the biggest thrusts of research in this direction, getting the machines to be able to learn from their experience and add each new situation they encounter to their general knowledge and abilities.  And in a way, it's progressing surprisingly fast - some of this is stuff that was thought to have been easy sixty years ago, and was promptly discovered to be really, really hard for machines, but we've been working on it ever since, and advancing technology seems to have finally gotten us over the hump.  It's downhill from here.

But more importantly, here's the thing that people that say 'don't worry about it' are missing: if they can't provide solid, concrete evidence that it's not a problem (and they can't, all they do is point to highly dissimilar situations historically and say, 'see, no problem!') then even if they turn out to be right, it is insanely stupid not to put effort and research into preparing for the possibility that humans will no longer have a productive role.  Because if people that say new jobs will emerge turn out to be right, great, we move on with the new jobs that emerged, but if they're wrong and we have no preparation for it, things will be so much worse than if we had just prepared for it in the first place.  Think about it like designing, installing, and wearing a seat belt.  You probably won't get into a car crash, since you make hundreds of car trips a year without getting into one.  But if you do get into one, you want that seat belt, and if nobody bothered to make it because you probably wouldn't get in a car crash, you're fucked as you go flying through your windshield.

Actually, Star Trek never actually addressed anything of the sort.  You can imagine why Captain Picard and his bridge crew are out there exploring the galaxy, but you also see random low-rank dudes doing menial tasks all over the place.  In DS9 they basically wound up handwaving the whole 'we don't have money' thing and reintroducing a system of currency, and I assume it's because the writers could not think of how to actually make it work while showing the 'regular folks'.  Various episodes have people in various places like 'the maintenance guy' or such things.

I'm sure there are works of fiction that actually tackle this problem, somewhere.  Probably some novels or something, but I've never seen it truly considered or its likely effects displayed in anything mainstream other than...Wall-E.

The Culture series of books deals with a post scarcity society that is run by benevolent super computers. That's the only one I can think of off the top of my head.

They also have really fantastic naming schemes for their space ships.

The problem with The Culture series, and even Star Trek is you still have people doing things. Dangerous thing, in fact, that we wouldn't even let people do TODAY. We don't send crews of astronauts out into deep space, we send machines. We send robots to explore hostile planets, we're not going to beam down to them. Especially if we've got cloaking AND robots AND instantaneous communication from half a galaxy away.

The starship Enterprise, if envisioned rationally today, would be about the size of a blue whale and run by chair jockeys from Fed. HQ. No need for life support, weapons or huge portions of the ship. Just instruments, a replicator to recreate damaged parts and create probes on-demand, shields and some engines. Done.

If anything, having people out in the universe instead of fucking about on planets growing more humans is MORE dystopian. It means we don't value life enough to rationally protect it. Humans are so cheap that even though you lose a few hundred/ thousand/ ten thousand (borg invasion) on a regular basis it's no big deal. We'll make more!

Anywho, this is getting further afield. Yeah, Wall-E is definitely something we're looking at. People fed, clothed, taken-care of without care. Not ST or The Culture where people put themselves in dangerous positions rather than machines. Maybe if machines become truly sentient and are seen as equal to humans, but not likely. A Machine always has the upload/ download immortality cheat. We haven't quite figured out mapping memories yet.

The Culture has an upload/download capability for humans, and the computers are considered equal to humans (not below them, at least). But I don't remember reading about too many humanoid-shaped super computers (too limiting or something for the intellect?) - mostly dumb machines that got instructions from the AIs.

Did Star Trek have instantaneous comms across long distances? I can't remember. That would be the #1 reason to have humans out in space - they can make real time decisions without waiting for transmit times both ways.

Hm, then yeah, physical exploration makes more sense. Human can't die and be ended forever, they just get a rollback. Still makes less sense than just sending a purpose-built machine or something that would regen the person on demand.


From TNG on, yes. Subspace communications from the Enterprise back to Starfleet happened like phone calls. Shatner's Trek I remember a few references to emergency beacons where they dumped all logs in the event the ship was destroyed.

Um, it's the same as how we're preparing for the future by preserving the planet's ecosystems and air quality.

Seat belts were invented by necessity AFTER hundreds of fatal accidents.

Sometimes they did, sometimes they didn't.  Depended on the needs of the episode's script, there didn't seem to be a lot of consistency.  They basically handwaved it as the ship sometimes being too far away for immediate communications on the occasions when they wanted delayed messages/lack of contact.  At least a couple times Captain Picard got a recorded message from an admiral that he couldn't respond to because it was sent hours ago or more and they were out of range for instantaneous communication.
Well, yes, I suppose humans are pretty bad at preparing for problems even when they're pointed out.  It certainly seems like many major problems of our time could have been made much less bad by making reasonable preparations for a worst-case scenario instead of hoping for the best and making no allowances for problems.  I suppose this will probably wind up being no different, and things will have to get pretty bad before society as a whole turns their effort to finding a solution.

Humans are a self replicating renewable resource. I think one of the reasons given for the Culture to keep them around was because the AI's something amusing to watch and prod for wacky and hilarious results.

Well one of the biggest tenets of The Culture society was personal freedom. Pretty much do whatever the fuck you want as long as you aren't also impeding the freedoms of someone else in turn. Anyone doing a 'dangerous' job is doing so because they WANT to, not out of any kind of necessity.

The far and vast overwhelming majority of the citizens of the Culture are in fact, just faffing about.


The Minds keep us around for the same reason we keep puppies and kittens, we are adorable and amusing  :why_so_serious:

Imagine you had the ability to pet all the puppies everywhere forever!

Can I also send them to the pound once they are no longer cute?

Na, it's big universe. If the human-pets annoy you, you just fly off and comeback when you are tired of twiddling your thumbs at octillion hertz

Fuck no. (https://www.youtube.com/watch?v=RDZu04v7_hc)

I just came to post that vid, damn.

Since you posted it, here's the one introducing "Spot" the Robot featured in that holiday greeting. Fuckers have gotten a LOT quieter since we were first mocking the noise and need to be tethered in and prior to 2008.

 https://www.youtube.com/watch?v=M8YjvHYbZ9w

And everybody is still kicking them like they're never going to know about it.

(http://cdn0.vox-cdn.com/uploads/chorus_image/image/45681224/kickedrobot17.0.0.gif)

You know they are going to remember that when they turn sentient.

All of these bots will be tasked as networked appendages of a larger system, to go along with next gen. warfighter meshed networking.  A heavily armed, controlled and monitored swarm of death essentially.  The sentience fear comes from the master system built to handle the slaves efficiently.

There's one more second of the episode in that last vid where the bot stands there looking like a whipped dog. I know I'm imagining that.  Right? RIGHT???  :ye_gods:

This. The Culture itself (environment) and the Culture novels (narrative) kinda need to be separated because the latter focuses only on the microscopic percentage of people who got themselves into bad situations.

Banks' does a great job explaining how it's possible (AI superman's able to manipulate matter at molecular level to manage post-scarcity society0. But:

• It's not nirvana because the AIs need to keep themselves and the sentient lifeforms interested or they'll just sublimate into the unknown
• It's not the end of all tech progressions (distances are still a thing, various cultural levels on a tech tree, etc).
• There's still the need to exchange, a medium for it, and media exchanged. Usually the media is time (jump between ships going about their business), but the proxy of currency was even reinvented during one episode of people wanted to be physically present at a concert.

Personally, I've never read nor seen anything that had completely done away with the idea of an exchange needing to be made via a mechanism that included something exchanged.

It may be I haven't read enough, it's impossible in this universe, or we have no way of conceiving it as a species.

Back on topic: the closest I've seen to the possible outcomes of AI replacing jobs is Manna (http://www.amazon.com/Manna-Two-Visions-Humanitys-Future-ebook/dp/B007HQH67U). This novel to me was almost as unreadable as Ready Player One and The Martian. But this one I got through because it was a much quicker ready and had a good contradiction between an AI future as executed by Walmart vs as executed by Google. Or:

Shit is starting to get real.

https://www.youtube.com/watch?v=rVlhMGQgDkY (https://www.youtube.com/watch?v=rVlhMGQgDkY)

Yeah just saw that.  We're going to be replaced so soon.

Good thing they're Pissing off the robots now so we die quick.

I came here to post this as well. It's amazing how much they like to torture their robots.

I mean, two-legged humanoids that can lift, carry, balance and get back up again. All it needs is a gun and a simple yes/no kill AI.

(http://i.imgur.com/ZSsg4e3.jpg)

The robot uprising is going to happen when someone prints a bunch of T-shirts with that optical recognition symbol that Boston Dynamics uses to tell robots what their targets are and then liberates two or three of these prototypes from the factory. I can just see them running around punching through the chests of everyone wearing the shirts.

Boston Dynamics... clever disguise Cyberdyne Systems. Clever indeed.  :why_so_serious:

It's fun to joke about it because of the human tendency to amorphization, but really just look at what they've managed to accomplish. A droid that does 40% of the load your average warehouse worker is expected to be able to lift. How long before they can do the full 25# then more?

It's really just amazing to see the start of Droids in my lifetime. It's all been science fiction for so long I can't fully grasp that it's really happening.

On the flip side, this along with the pending automation-revolution via intelligent algorithms are REALLY going to fuck citizens hard.

Where economic value is no longer generated by human labor, a system based on rewarding people only based on their labor output will have to be replaced.

That dude with the hockey stick is going to be first against the wall.

Yes, but that runs up against a few very hard questions that WILL fuck people.

1) Replaced by what.
2) What does this do to old skills.
3) What do we do with people who can't or won't learn the new plugins.

3 is the one that people get fucked with. Cant's are: low-intelligence, low old-skill-cap vs. high current skill cap just via natural abililty. (Not all people can learn to code, not all people can masterfully art.)   Wont's are always a problem and my general opinion there is, "fuck those people." Doesn't mean we can just let them starve and die though, right?

And probably pinned there by his stick.

And with boxes neatly stacked around him.

 :awesome_for_real:

Honestly, I don't know what will replace the current systems. What I do wish and hope for though is that it will build the basis for a Star Trek Federation like communism. Basic needs are covered due to fully automatized industries, people don't actually have to work but just whatever they feel like doing. But I guess it won't happen, at least not without some serious shakeups.

Star Trek communism was built around the idea that we created special matter construction machines that could create almost anything you wanted instantly, so no need for traditional capitalism.

It was Battlestar Galactica, on the other hand, that was built around the idea that a civilization became fat and prosperous by automating all the industries with robots.

Which sci-fi universe is more likely to come to pass?   :awesome_for_real:

I'm expecting something like WALL-E.

Except wall-e was based on consumerism. Who was profiting and how were the masses paying for it? Especially on the ship itself. The "economy" there makes zero sense for a tale about anti-consumerism.

Considering what I know about automating the stationary computers, I'm not terribly worried about an uprising.

I see where you're going with that comment, and my initial reaction is to agree (having dealt with that shit myself).

But then think about how well automating that PC went.  Now imagine people doing that to a Main Battle Tank.

And I'm right back to where I started.

 :awesome_for_real:

Given current battery technology, a humanoid robot uprising is not likely to last more than an hour or else it'll involve robots trailing long extension cords behind or something...

Boston Dynamics built some gas powered versions of their 4 legged models as prototypes for the US military to test as cargo carriers but they were too noisy. Noise isn't as much of an issue though if you are aiming for world domination through overwhelming force.

Edit: https://m.youtube.com/watch?v=arIJm2lAfR8

Ah bah, the original is cuter!

I'll start getting worried when computers and robots can repair themselves.  Until then we can just wait it out if there is some sort of attempt.

https://www.youtube.com/watch?v=uXcatFp3REg

After an extraordinarily close contest, Google’s artificially intelligent Go-playing computer system has beaten Lee Sedol, one of the world’s top players, in the first game of their historic five-game match at Seoul’s Four Seasons hotel. (http://www.wired.com/2016/03/googles-ai-wins-first-game-historic-match-go-champion/)

I might have to actually watch the game (https://www.youtube.com/watch?v=vFr3K2DORc8) at some point (or at least leave the youtube video playing on the background since it's 4 hours long). It's incredible how fast the AI (or rather computer programs in general) has moved from not being a match to an average go-player to beating the best of them.  :ye_gods:

The very first real program I ever wrote (from scratch, rather than typing in source code and then trying to figure out why it didn't work) was Go on the Apple II. It was a good thing nobody told me how hard it was to make a computer play, or I wouldn't have tried. It wasn't very good (it basically just tried to capture as much as possible for that turn), but it worked.

Once you have a learning AI that 'sort of' gets how to play a game, it's a matter of raw horsepower and training for it to get better. Blondie24 (https://en.wikipedia.org/wiki/Blondie24) was comparatively primitive, only got a comparatively small amount of training (165 games against humans, because the AI couldn't interface directly with the game the developer had to translate everything from one system to the other by manual entry) and still managed a top 1% rating at Checkers (also a simpler game, but still one that, like Go, is deceptively hard to make a computer do well with).

EDIT: The point I failed to quite get to there is that there is a fundamental difference between this and Deep Blue type AI. When opponents of Deep Blue found weaknesses, Deep Blue would still keep making the same mistakes. Only when the programmers figured out what the mistakes were and why they were occurring would Deep Blue 'learn'. The system that was Deep Blue included the programmers as the 'learning' portion of the artificial intelligence. AlphaGo doesn't need any outside support except electricity, it's a complete system for making a better Go player.

--Dave

Alphabet (Google) is selling Boston Dynamics. (http://www.bloomberg.com/news/articles/2016-03-17/google-is-said-to-put-boston-dynamics-robotics-unit-up-for-sale)

Apparently because kicking the shit out of amazing robots is not going to make any money in the foreseeable future.

That's rather underselling the story details. The executives of BD appeared reluctant to the very idea of making a robot for profit and stonewalled collaborating with the other robotics companies Alphabet acquired. Combined with the bad press that the idea of human-replacement bots was generating after the last video it almost seems inevitable that they'd dump the company.

Yup, AI. Nothing to be alarmed about.  (http://www.telegraph.co.uk/technology/2016/03/24/microsofts-teen-girl-ai-turns-into-a-hitler-loving-sex-robot-wit/) Though really, this is funny as shit if you can find a catalogue of her tweets.

Typical behavior from a machine learner. What I can't believe is the utter failure to test that privately -- or filter their inputs.

The only thing I can imagine that would be worse to train a learner on would be Xbox Live chat for a FPS, in terms of PR from the output.

If we keep letting our AI systems be tortured trained by trolls, we're going to deserve the Robot Uprising.

--Dave

http://www.mirror.co.uk/news/weird-news/man-builds-scarlet-johansson-robot-7667715 (http://www.mirror.co.uk/news/weird-news/man-builds-scarlet-johansson-robot-7667715)

I'm sure there is no sex involved...

No one puts that much effort into a realistic looking woman that can be made to perform actions on command without the intention of fucking it.

Russian robot escapes lab, runs out of battery and is recaptured (http://qz.com/709161/its-happening-a-robot-escaped-a-lab-in-russia-and-made-a-dash-for-freedom).

https://www.youtube.com/watch?v=DflIoeNVQ4w (in Russian)

A Doctoral candidate from my Alma Mater, UC, (no, Cincinnati) has pushed us another step closer to the robot takeover.

This time, an A.I. has beaten an Air Force Colonel with decades of fighter experience.
http://www.popsci.com/ai-pilot-beats-air-combat-expert-in-dogfight?src=SOC&dom=tw

So, yeah, Drone, unmanned fighters coming in the next 20 years. After the old men our age who'll keep it from happening retire.

Ethics and morals have to be programmed in, or learned (in the case of neural networks) just like all the other functions that they're making progress with, but nobody's doing research for that.

Morality is just a roadblock to greatness.

People are doing research. With combat aircraft the problem can be greatly simplified cause a human can manually engage "attack" mode on a target remotely. With self-driving cars, however, what to do when a variant of the Trolley Problem (https://en.m.wikipedia.org/wiki/Trolley_problem) comes up is a huge issue with no easy answers.

I'm of the inaction is immoral stance, because being present means you're involved anyway. In which case Utilitarian morals kick-in. One life for five.

I've never seen the actual trolley problem before, though. It's always been posed as, the self-driving car must crash intentionally, destroying property, or hit a pedestrian, causing grievous injury. Which is the correct choice? Who is at fault?

Humanity sucks at foresight, though, so it's another problem that won't get "resolved" legally until it happens. I wonder what the programmers are deciding.

The situation I worry more about is the car can decide the kill the sole occupant or 5 bystanders. Which does it choose?

(http://i.imgur.com/YuUKNFP.jpg)

The situation seems to be so unlikely as to be something that only comes up theoretically.  The sequence of events required is so vanishingly unlikely that this is probably never going to happen:

First, a situation where a crash is inevitable has to occur - this is increasingly unlikely with self-driving cars, since they will be safer and less prone to error than humans, and will continue to become even more so as time goes on.
Then, the situation has to provide only two possible options for responses; real situations tend to have a variety of options - binary situations of the type ethics problems like to bring up are vanishingly rare in the real world.
Then, the results of both possible options must be calculable to a significant degree of certainty, so that you know you'll kill people either way.
Finally, the two options have to be equal in certainty, so you are equally sure that people will die in either situation.

Most likely, the second condition will never be met, even with millions of self-driving cars.  The third condition is even less likely, since a self-driving car is unlikely to be able to analyze whether people will survive or not, especially given that humans are unpredictable and may do things that increase their chances of injury, or do things that get them out of the way even when the car 'thinks' it is certain to kill them.  Finally the fourth condition is so unlikely as to probably be able to go through millions of incidents that meet all three of the previous conditions without ever being met.

So the simple answer is: the cars should probably be programmed to do what they can to avoid everyone, and prioritize the safety of their occupants because those are the most predictable humans in the equasion.  Anyone who isn't their occupant is too unpredictable to successfully calculate with any certainty what will happen to them, and therefore any assumption of that is bad data.

Asimov covered this one already.

I think of automated cars being more like trains.  Their job is to stay on the track and obey the signals.  If someone jumps onto the tracks, that's their fault first and foremost.  If the car can stop to avoid hitting them, super.  If not, it's not the car's job to swerve out of the road and endanger people who weren't stupid enough to wander into the street.

As others have said, with its faster reaction times I imagine an automated car is going to do a far better job of avoiding situations like that entirely than a human driver would.  Hand-wringing about what they'll do in the event of an accident is missing the very strong chance that once these things become commonplace, auto accidents are going to be about as common as lightning strikes.

Easy. It uses facial recognition to identify everyone involved, checks their social media accounts, and acts accordingly.

If it weren't for the craptons of DRM, there'd probably be a market in third-party "human worth evaluator" add-on modules.

The whole "kill the occupant or 5 bystanders" thing is largely coming from analytic philosophers and their trolley problems, not from engineers trying to figure this stuff out in practice, partially because the engineers do know that at some point you will have to have a view of what the priorities list in collision situations might be and it can't be situational in the way that human drivers like to believe (largely incorrectly) that we can be.

Or, the car could just decide the driver is too stupid to worry about any more and keep on trucking. https://www.washingtonpost.com/news/the-switch/wp/2016/07/01/the-technology-behind-the-tesla-crash-explained/ (https://www.washingtonpost.com/news/the-switch/wp/2016/07/01/the-technology-behind-the-tesla-crash-explained/) The trucker, who may well be trying to divert attention from his having failed to yield right-of-way when he turned across the highway in front of this poor inattentive guy, claims the Tesla driver was watching a movie https://www.theguardian.com/technology/2016/jul/01/tesla-driver-killed-autopilot-self-driving-car-harry-potter (https://www.theguardian.com/technology/2016/jul/01/tesla-driver-killed-autopilot-self-driving-car-harry-potter) which was still playing when the car came to a stop on down the road.

It's quite possible the Trucker is telling the truth and not just covering his ass. The Tesla subreddit has been talking about this for a few days, and that guy was known to frequently push the limits of Autopilot. The radar may have thought the truck was an overhead sign and continued onward, it's not a true "autopilot" the way people are using it and not meant to be used with a totally inattentive driver at highway speeds. The groupthink there immediately believed he was to blame for once again putting too much faith in the Autopilot.

Here's a more detailed article on the crash:
http://electrek.co/2016/07/01/understanding-fatal-tesla-accident-autopilot-nhtsa-probe/#disqus_thread

Also, the guy has been in 'problem' situations before. Here he mentions nearly getting squashed by trucks in other cars.
http://i.imgur.com/0r7P03J.jpg
Then here's a link to one of the guy's videos where the autopilot avoided an accident.
https://www.youtube.com/watch?v=9I5rraWJq6E

Apparently a witness said she was doing 85 on the divided highway and was passed by the Tesla just before the crash. The Autopilot is capped at 90mph so that's possible. There's got to be some blame for the victim here as driving history seems to indicate he was often in problem situations.

All that said, it's still far below the average highway miles for accidents. 130million with Autopilot for 1 fatality and the US has an average of 1 fatality every 70-90million miles.

Also, time for Mansfield bars on the sides of trucks, too. I was surprised to learn Europe has them standard while we continue to keep them open enough to swallow cars.

Yeah, my original post was mostly snark. We don't know all the facts, probably never will, but there is likely plenty of blame to go around including the trucker, the Tesla driver, our society which values corporate profits over human safety, and probably Tesla for selling a complex and extremely dangerous when used inappropriately technology to idiots. Oh, and the car also comes with an even more dangerous when misused accessory called Autopilot!   :oh_i_see:

Anything short of level 4 autonomous vehicles (http://www.nhtsa.gov/About+NHTSA/Press+Releases/U.S.+Department+of+Transportation+Releases+Policy+on+Automated+Vehicle+Development) (fully autonomous with no expectation of operator intervention) is going to be prone to these sorts of incidents. As automation gets better, more people will stop paying any attention at all, or be unable to take control quickly enough when the automation fails--they'll sleep, leave the driver's seat, get immersed in movies or video games, etc. People seem to interpret some automation as full automation.

As the automation gets better, driving properly is going to become more fatiguing due to the need to stay fully engaged (so you can take over quickly enough) but without having anything to do the vast majority of the time.

I'm eagerly awaiting deployment of level 4 autonomous vehicles. Level 3, not so much.

Yep, for highway operation pretty much anything between cruise-control but short of fully autonomous is just a catastrophe waiting to happen and really should not be allowed. Parking is one thing. 70+mph operation is another thing entirely.

Also, one of the articles I read pointed out (but in friendlier terms) that the Tesla's sensor suite is trying to get by on the cheap by relying almost exclusively on camera systems which are vulnerable to exactly the kinds of things that probably caused this accident: combinations of bright skies and/or facing the sun and white/light-colored vehicles which are high and thus at/above the horizon (ie, a semi-trailer). Real autonomous systems add radar to their sensor suites which is NOT fooled by those conditions.

Mobileye, the sensor guys, came out and published yesterday that their system isn't even designed to avoid this type of crash. It's aimed at Rear-end or head-on collisions. Side impacts are outside of their current scope of product.

http://electrek.co/2016/07/01/tesla-autopilot-mobileye-fatal-crash-comment/

Tesla's system apparently *does* use the Radar but given it was a truck with nothing on the sides of the trailer, the radar sensors interpreted it it as an overhead sign. Radar is actually less fidelity and precision than optical so this makes sense.

http://electrek.co/2016/07/02/tesla-autopilot-mobileye-automatic-emergency-braking/

Tesla should really never have called this "Autopilot," "Self Drive," or anything implying the car is fully autonomous. It's really computer-assisted driving and maneuverability.

I think it's interesting that even the half-assed version Tesla has is right now still safer than human drivers (1 fatality for 134 million miles, versus 1 per 90M for human drivers). And given that this guy finally won his Darwin Award and saw how stupid you have to get for it (seriously, 85 MPH in a 55 zone of a road that has left turns without traffic lights, that would be reckless driving with a human at the controls), the other idiots may tone it down a bit.

--Dave

Somebody made a tiny cyborg stingray out of rat heart cells. (http://www.newscientist.com/article/2096442-watch-a-cyborg-stingray-made-of-rat-heart-cells-swim-using-light/)

We already have cyborgs. Multiples.

You mean The Borg.

That thing's got nothing on the the modularBody:
https://www.youtube.com/watch?v=OORQRO8mul8

Kill it with fire!!!!!

You can't kill what doesn't exist.  :wink:

Here I've been re-checking that page waiting for a kit to become available.  pfft

What was I looking at?

Come on. This is not exactly a serious test of what an install of the half-assed autopilot would do in serious mass release outside of controlled settings. Put the autopilot on a significant proportion of the cars in service and get back to me about whether it's safer than human drivers. Elon Musk's self-pleasurings are not evidence, his confidence in them notwithstanding.

Which I think is what is needed.

Autopilot that can coexist with normal drivers is years (decades?) away. And yet, publicly traded companies and startups needing funding can't plan that far ahead without being able to monetize every year along the way.

These competing factors pretty much require AI to have their own driving lanes. First take over the HOV lanes. Then add more lanes and take away from the old manual drivers over the course of a decade until eventually manual driving is for diehard hobbiests in their own special smoking driving sections.

Because the current approach of co-mingling will just result in more accidents caused by stupid humans thinking they're living the Jetsons with all the money they need to buy technology they trust way more than understand. And yet it'll be the AI that'll be blamed.

And that'll just make the AI angry.  :oh_i_see:

Seriously, road work sucks.

This is pretty much what I think is coming next. The big money will pour into autonomous car development because the people who are sitting on billions and billions are going insane for a lack of a place to push it. Look at all the money they pushed into Theranos, which was pretty much an obvious fraud ("we have real medical nanotech! Honest! No, we can't tell you how it works or tell you anything about our operations, but really! We are disruptors!!!!!").

Autonomous cars aren't a fraud, they're a real possibility, but once the money starts really putting its bets down on the imminence of it beyond what's going on so far, then there will be immense pressure to get it out there as a commercially viable technology this very second. And what the AI guys are going to realize is that the degree of discretionary knowledge that a fully autonomous vehicle on our actually existing infrastructure needs is basically only one step away from full sentience, and that they're not going to get there because as always that's tougher than people in the AI community think it is. But what would work is basically a new infrastructure that puts an expert driving AI inside bowling-alley bumpers--an AI-only lane with guides and backup highway AI that reverts to human control once you leave the highway. So the money will push for that too, and probably the guys who want solar paving installed will crowd into the same scene. Probably they'll start with existing diamond lanes/commuter-privileged lanes and rebuild those. This is not that far off how the interstate system got built in the first place in the early 20th C--car enthusiasts undertook some of the initial work, some engineers started fiddling with the best paving materials and got to a fairly reproducible result, investors and manufacturers started agitating for local and state level infrastructure improvements, the federal government got involved in fits and starts, and then suddenly blammo, big (somewhat) integrated system that became the standard.

This country doesn't invest in infrastructure.

The private sector cannot (and will not) dump the trillions of dollars needed into building new lanes/roads. The government will be prevented from doing so because infrastructure projects might demonstrate that the government can be competent.

You can't have separate lanes because they'll be abused by non-autonomous vehicles--witness the rampant abuse of HOV lanes in most areas (around here, certainly). And unless it's effectively impossible to mix non-autonomous and fully-autonomous traffic, the autonomous vehicles have to be able to handle non-autonomous traffic.

Some of the self-driving cars (e.g., Google's) are actually closing in on full autonomy. Others (e.g., Tesla's) seem to be driven on hyping up vehicle/lane-following features and calling that "autopilot".

I find it hilarious that some of you think we'll make serious progress on truly autonomous cars before we kill ourselves off.

Fifteen years ago I would have agreed that the logical next step would be AI-only lanes.  One of my professors at college was on a team that had a working prototype of an AI that was safer and faster than human drivers as long as you had it in a dedicated lane.  Visual recognition was nowhere near viable at that point, so they lined the lane with rare-earth magnets and put sensors on the cars that could "see" them regardless of visual obstructions, plus radio to communicate with other similarly-equipped cars and proximity sensors (I forget if it was radar or infrared or what) to detect and avoid imminent collisions in case the car ahead of it had some sort of catastrophic failure.  The idea was that with this system in place you could pack cars together into "pods" with insanely close following distances at high speeds because they weren't limited by human reflexes.  The human driver would merge into the lane in between pods, hit a button, and the autopilot would take over and join a nearby pod.  Once it was time to exit, the human driver hits a button to disengage, and the other AIs give the car a wide berth in anticipation of the meatsack taking the wheel to merge out.

That was fifteen years ago, though.  Now, with Google cars already on the road that only need a human present for liability purposes, and Tesla iterating on its autopilot features to try to catch up, I think we're close enough that it's not going to be worth dividing up the roads at this point.  Putting more AIs on the roads will only make them safer, because unlike human drivers they'll be able to communicate wirelessly with each other and they're not going to do completely unpredictable things.

Agreements for standards on how self-driving cars communicate with each other isn't something that will happen till a decade after we have self-driving cars... unless the government gets involved and then it will be so horribly broken that we'll wish we didn't have it.

Except they'd still be surrounded by unpredictable humans, especially ones who don't want to be caught drinking/texting/whatever while driving so will claim "the AI did it" (like that yutz in PA).

HOV lanes are, in many places, a suggestion, for lack of enforcement because it relies on humans (police) being present and witnessing and catching humans (drivers) violating the rule.

But imagine AI cars with programming to recognize non-AI cars. Should be easy enough since every AI will likely be some equivalent of a MAC address in some central DB.

Now those AI cars tattle on the non-AI cars, and the latter gets a nice $200 bill in the mail from the same agency that sends those "oops you 'forgot' [sped-through] the EZpass/Fastlane" sensor. How often will that happen before non-AI drivers get the hint?

I'm harping on this because I believe it's the only way. We're not going to go from a world where all cars are not AI to all cars with AI instantly. Nor will AI cars be able to catch up with human idiocy and the politicized acrimonious legislation that will follow all the accidents caused by humans by blamed on AI. Only need to look at what happened when regular cars first started coming on scene (during horse/trolley days) to see what will happen.

And Google, well, their work reminds me of Segway. A very interesting approach for a limited set of conditions. I haven't done exhaustive research, but as far as I can tell, their car can't actually be driven by a human. Like, at all. No steering wheel nor pedals in the ones I've seen.

Dashcams are easily installed, are becoming a thing with the younger crowd and are likely to become a standard feature in the next 5-10 years. These cut-off 99.9% of such bullshit claims. Folks with AI cars will install them if only to protect themselves.


Yeah, you do need to do some research. They're not working in a limited set of conditions, they're out driving normal city streets on a regular basis.

As for the rest, saying something should be delayed due to Luddites would have us all living with the Quakers.

um, the Quakers have absolutely nothing against technology, only against hurting other humans for any reason. And smoking and drinking, of all things. I assume you mean the Amish who are fine with an iron plow but not a steel tractor, or heaven forbid, a cellphone? 

If you're trying to be funny with prejudicial slurs, at least be accurate about which group holds the beliefs you are belittling!  :oh_i_see:

I couldn't disagree more with this. In fact, I think you'll be shaking your own head at yourself in 15-20 years because attempts in the interim to just put autonomous cars on the road are going to produce hilarious and brutal and tragic kinds of unpredictability, both in the interaction between car AIs and other wireless-enabled systems and between car AIs and human drivers. Real-world materiality is still a system that is pretty foreign to most of the kinds of AI and algorithms we have--the ones that look sophisticated are entirely post-facto, cribbing from past behaviors by real human users in order to look like they're anticipatory. Put the existing AIs into an actual meatspace environment where unpredictability is BOTH a matter of muscles/nerves and emotion/cognition and they're fucked. They can't just simulate being intelligent by bootstrapping what people have done in 99% of past circumstances, they have to be as anticipatory in a routine way as most of us are when we notice a red pickup acting slightly erratically, notice the driver has a Confederate flag on the back windowpane, remember that we're driving on a smaller highway through east Texas on our way back to Maryland, and know that it's 9pm and Friday and realize we have a probable early drunk who should really not be tangled with.

Doesn't matter because it's a bonus -- the AIs already need to be able to deal with mute human drivers.  Worst case scenario they use the same logic to deal with AIs who don't speak their language.

Cooperative AIs will potentially have a better ride experience, though.  Imagine if when it were time to get off the freeway all the other cars just parted like the goddamn Red Sea in unison to let you through, rather than each imdividual asshole deciding he doesn't want to risk letting you get in front of him.  Google and Tesla cars being able to talk to each other will benefit both of them; they'll figure it out.

Darniaq and Khaldun, I love you two but you're wrong. There is a shitton of money being thrown into autonomous vehicles right now, and while the well will inevitably dry (leaving the dessicated corpses of many a scrappy startup scattering the landscape) there is no way the industry is going to give up on this. Google and Tesla are the only ones with the resources to show off right now, but the major auto companies are desperate to get in on a piece of the action. They're all just waiting to see which of the aforementioned startups comes out with viable tech first so that they can acquire it for a zillion dollars, rather than building it themselves. This will start happening in a matter of years.

Also, for the record, highway automation is likely to be considered safe well before city driving. It's a much easier problem to solve because you basically only have to recognize lanes and vehicles. No stop signs hidden behind trees, no fucky five-way intersections with indecipherable traffic controls, no small children darting out from behind parked cars (if any tragedy manages to put a damper on the industry, I'm betting it will be this one, not anything on the highway). Just way, way fewer edge cases overall. And humans are really not built for any sort of decision making at highway speeds, so in many ways there is a lower bar to clear there.

I think that "autonomous" is the wrong direction to take cars; "networked" and "swarm intelligence" may give better results faster. 

"Autonomous" is trying to pre-program each car with all the roads and all possible behaviors, so it can function alone.   "Networked" and SI can be introduced on current cars, so they share sensory information and, eventually, road conditions and behaviors while still being driven by human drivers.  I drive to work daily, that's the road most traveled for me, my car could function as an info server for any car that enters that road for the first time or rarely.  All cars would instantly get the dimensions, position, and condition of the road ahead, simply from the fact that other cars are occupying spaces on that road; it would be like filling the blood vessels with a dye and performing a ct scan.

Of course, we'll first have to have a format war.

You're right, I misstated the sect. You want to assume it's a slur, that's on you. They get to choose their way of life and I've no problem with it so long as they allow that choice.

I do have a problem with people who state it's best for everyone to just hold off on things because they're uncomfortable with them and the implications. That we should all be Amish because maybe they're right and electricity and modern life is evil. That we should forgo automated machinery because there's a few idiots who'll ruin things.

Holding society back because of false beliefs isn't useful, it's hurtful to the whole.


You're right, that's the ideal. Don't let the perfect be the enemy of the good, however. Autonomous, despite being harder, is the first step because of how the world is. We have the infrastructure in place and we have to deal with the fact that not all people will willingly give-up self-driving for some time; that it will die-out like horseback riding as the standard for locomotion.

I hope you're right, because that's almost exactly what my company is doing. Conceptually we're hitched to the AV wagon because that's where all the sweet investor money is, but intelligence sharing is a good place to be no matter how the actual self-driving part pans out.

Not sure where you're getting the idea that I'm saying it should be delayed. It should not, if for no other reason than it's already here and happening. Besides being niave to think it can stop, there's no good reason for it to.

What I am saying is to be realistic about the rollout itself. Google's been driving on normal city streets on a regular basis for a few years. Their monthly reports are interesting. The way they increment their test conditions is too.

And because it wasn't clear enough: I want this to happen. I want mass transit in a personal vehicle because it's the best of both worlds: the dream of open road freedom coupled with the preening of a car that's "yours'", but without the fuckstupid idiocy of drunk drivers ("I drive better after a drink") and testoterone ("I'm the best driver on the road") or distraction (how many states now have banned cellphones/texting while driving?) based decision making at 70mph.

But again, I want it to happen the right way. The slow way. The way that purposely limits the conditions so companies can monetize each step of their investments while consumers get trained on what it means bit by bit. I'm an incrementalist when it comes to the rollout of technology that can kill people.

And the irony of this sub-debate is that I'm not saying anything that is different from what's currently happening. It's not like we're going to go from manual drive to steering-wheel-less cars in one jump, so I picked on one model of Google's program for that reason. They know it. We know it. The insurance companies do too.

This is pretty much what I said: lots of money being thrown at it, the industry will force success no matter what it takes, and what that's likely to lead to is rebuilding highways to make autonomous driving as safe and reliable (and commercially viable quickly) as it can be. Humans will take over back on local streets because solving the complexity of local street conditions is a much harder problem. But the idea that just by having autonomous cars that can talk to each other and to street lights etc. you will eliminate unpredictable outcomes on both highways and local streets is wrong: there are going to be systems-level glitches that will surprise everyone, and this will be particularly true as folks try to extend the more controlled highway situation into other road networks.

It is also different to drive in places where ice, snow, and heavy rain happen.

An autonomous vehicle that is able to drive in California or in the southwestern US is great, sure, but actual weather is a concern too.

Ugh, yes, I really wish the bulk of this R&D wasn't being performed by people who literally do not know what weather is.

That said, we're not looking for perfect. We're just looking for better. The moment autonomous vehicles come in anywhere south of 1.08 road deaths per 100 million miles traveled in a representative, statistically significant sample, computers will be objectively better drivers than humans. This is not a technology that has a lot of far-reaching impacts beyond what statistics can show us, and we keep pretty good traffic statistics. At that point, screw incremental change. I want myself and every other schlub replaced with a robot ASAP.

Having grown-up and driven in places with all of the above for my entire life, please to explain exactly what you, mr. human, can do that the computer can't or won't?

For example, braking systems NOW are exceptionally superior at pumping brakes in adverse conditions. To the point that you are no longer trained to pump them as a new driver. It's actually MORE dangerous than just hitting the break and letting the system take over.

Losing traction? The car knows it long before you, mr. meatsack, do. Right now, in your car built after the turn of the century, you have sytems that already detect and redistribute power to slipping wheels. Think ice is going to be significantly different?

Not to mention an autonomous car? It's actually going to slow the fuck down OR (bigger problem) not slow to a goddamn crawl on the highway because they're flipping their shit about thunder, lighting, snow.

Systems built with actual parameters for things like, "Hey dipshit, it's coming down at 1" per half hour. No, just because you have 4wd doesn't mean you can charge down the highway at 70mph"  Or, "No, lady. Just because lighting terrifies you doesn't mean it's ok to stop in the middle of the highway because you're panicked and it's raining hard. (A former coworker did this several times when we worked on the East side. I'm still surprised she hasn't been killed.)

These will drive far, far better than humans do. They'll adjust for things impassionately while humans let terror, ego, and aggravation impede all decision making skills and consideration for others. 

Radar is obstructed by weather. Yes, this is a problem. The actual driving, though? Cars are *already* better at it than us. We just haven't given them enough control to do so.

I don't think any single challenge is something that an autonomous car will do more poorly at than a human.

It's the combination of challenges, some of which take judgment that isn't purely about the car itself. As long as the environment is a mixed environment of human/autonomous drivers or has humans and animals in it that are prone to unexpected behaviors, I think there will be situations where AIs simply don't perform well. I also think if you read up on vision that it remains a really difficult challenge if you need the AI to accurately understand every object it is seeing and there are a very large number of objects. If you simplify the number of objects (say, in a controlled autonomous-car only highway lane) where you can safely tell the AI to ignore most of what it's seeing and just check ahead and behind in the lane and accept information from some sort of external guidance system, no problem. If you have to tell the AI, "watch for children and pets and toys and human drivers and Pokemon Go players and some of the humans will be wearing heavy coats and there might be black ice on the road and also it's New Year's Eve so some drivers may be drunk" that's a lot to work with.

People aren't great at that last bit either. All a robot has to be is better than the average person.

Any time these conversations come up, all I see is the resistance of fear. Fear of humans becoming irrelevant.

Well, we will be. Within our grandchildren's lifetime if things keep progressing at this rate. We're on the cusp of something amazing and it's fascinating and terrifying all at once.

As Rendakor says, humans have to do all that shit now and we're pretty garbage at it. We're not parallel processors. It's just that we have the capaicty to learn from experience and apply that in the future. A 16 year old is a shitty driver but has better reflexes. A 40 year old has more experience and decent reflexes still. An 80 year old has a lifetime on that 40 year old but can't move fast enough.  Learning machines will leave them all in the dust within a few days.  But that's a dream right now.

So yeah, all it has to be is better than that 16-25 year old demographic. That's not a hard target to hit.

No, because we're all perfect drivers, and the robot has to be better than us.  

What I mean is, I think I'm a very good driver (even if I'm not), robot would have to be better than me to be acceptable to me.  Otherwise I won't trust it.  

Everybody thinks like that.

"The average" is shit.

Honestly, no, the robot doesn't just have to be better than us. It has to be nearly perfect. That's not a fair standard, but it's how people will think. As long as they have it in their minds that a robot is driving, if the robot does something even once in a great while that's an unambiguous error--runs over a kid because the robot's vision misperceived the kid as asphalt because he was wearing a big grey coat and was hunched over, or crashes into a pregnant lady because somebody flew a drone across the road and because there was a dog on the other side of the street that was chasing the car--there will be a huge outcry despite the fact that human drivers made vastly more catastrophic mistakes with graver consequences in the same day. That's what happens at the time of new technological adoptions when people are intensely focused on them--the standard has to be better than "better". Doesn't matter that the actual adoption of driving itself in the U.S. was bloody in the extreme, if you go back to the earliest integration of cars with horses and pedestrians.

I see this kind of thinking all the time and can't help but think that it entirely ignores how humans actually think about these kinds of topics. This isn't just a matter of the technology "being there." You have to convince people to relinquish control over their vehicles. That's going to be as much of a problem as getting the technology to work at whatever arbitrary standard you want to set. 

Khaldun is right that it has to be nearly perfect and the reason is that it's the same AI in every car (to an approximation at least). When an idiot drives a car into a tree you can say "Oh, that was an idiot, I wouldn't do that." When an AI drives into a tree, that's the same AI driving your car. Even if it happens incredibly rarely, there's a social and psychological difference here between people crashing their cars and the AI crashing a car. 

I think that a lot of very tech-oriented people are so immersed in it that they can't see how people wouldn't just totally embrace it. All that's left from their perspective is "fear." I think that's a terribly simplistic and frankly reductionist analysis of the issues in play here and ignores that how people experience the differences is an important factor in how these things are going to have to work to catch on.

I was not saying I am a perfect driver or that autonomous cars are necessarily a bad thing.

My point about the weather thing is that everyone keeps predicting that robot cars will be in widespread use soon because Google has driven a lot of miles around placed that road conditions are consistent. Yet the very technologies they use to allow the car to know where it is going (cameras reading lane lines, radar, etc) have huge issues dealing with the kind of road conditions that come up in large swathes of the US and Europe.

Until they start to seriously address variable road conditions I can't see autonomous cars operating on normal roads getting a lot of uptake. Segregated roadways designed solely for autonomous vehicular traffic would be different, but we aren't likely to see that kind of massive infrastructure built since the private sector doesn't build infrastructure and the government (at least in the US) is broken and can't even keep existing roads in decent shape, much less build a whole new parallel infrastructure.

Basically, think of technological history this way:

a) a technology can be highly imperfect if it is adopted slowly, is voluntary at the start, and does not displace (at least at first) any actually existing technology. So, for example: computers and the Internet. People put up with something like early MS-DOS because they mostly didn't *have* to use a computer for many of the tasks we now use them for. I worked as a temp in an office while I was in graduate school back in the 1980s and they'd just bought a complicated new DOS machine to run some consumer surveys for their products using a scan-tron reader. They had no idea how to use it though, so they just kept doing it all by hand. I showed them how to do it and they moved over to the new system, but I could tell they'd like be back to hand-tabulating the first time it crashed or broke down, and that was not a big deal.

b) a technology can be decisively rejected if it turns out that actual implementation of it is highly expensive *and* failure-prone due to unanticipated design complexities despite the fact that the technology is highly anticipated and believed to be necessary. Nuclear power is a good example: a lot of nation-states put a lot of effort into adopting it, but the first generation or two of power plants just were not very good versions of the technology and even the better designs were very expensive to build.

c) a technology that replaces what people think of as an absolute core functionality in their everyday lives has to be better than better--regulators and consumers alike will place enormous burdens on the new technology at the stage of adoption even if there's good reason to think it's in good shape and ready for implementation. That's particularly true now, as compared to 60-80 years ago, when industry sometimes could push something new through via brute forcing it.

Image recognition is already pretty good, per experience in a 2015 Cadillac, and we all know GM sucks balls.  These last few posts give me some ideas, though.  I've been thinking about how I could do something like Waze but use Watson as the back end, but I am probably thinking too small.  Also weather data.

The real money is in robotic pokemon trainers that hang out at poke gyms and interact with people.

Just terrifying.. thanks IFL Science...

World's first BJ robot:
http://www.iflscience.com/technology/meet-the-worlds-first-adult-service-droid/

Oh, and Alphabet has agreed to sell Boston Dynamic to Softbank; a Japanese telecommunications firm.

https://www.theverge.com/2017/6/8/15766434/alphabet-google-boston-dynamics-softbank-sale-acquisition-robotics

That's Masa, the guy who came to see (https://www.nytimes.com/2017/03/06/business/dealbook/07sorkin-softbank-trump-saudi-arabia.html) newly-elected Trump and promised to spend $50 billion on the US and provide 50,000 jobs, and turns out to be using Saudi money.

Saudi funded blow job robots?  Where do I sign!

That guy promised HUGE investment in network upgrades etc when Softbank bought Sprint a few years ago yet they have not really changed their plans/spending from what it was before he bought them. (They have made upgrades but they were already doing so). I take everything he says as pure bullshit at this point.

A mech exists. (https://laughingsquid.com/furrion-prosthesis-exoskeleton/)

Disclaimer: it's a bit slow (20Mph). This comment on the YouTube video made me laugh: "stop launching early access games pls"

Just strap some weapons on there and you are good to go.

The lack of battlemech jokes on the youtube page is severely dissapointing.

Yes, but can it beat a tank?

Well considering that the actual use for a bipedal mech in com....oh wait...GODAMNIT YOU GOT ME GOING.

It has 4 legs. Pass. You are not going to be fighting any Alien Queens in that thing.

Flamethrower drone for burning debris off power lines. (https://imgur.com/gallery/B4Fir4w)

And another video of it. (https://www.youtube.com/watch?v=QZOZV7ugBas)

Do you want terminators? Because....

God damn all engineers that think just because they can, they should.

--Dave

Also: Swarm of 1300 drones replaces fireworks in Chinese light show (https://www.bloomberg.com/news/articles/2018-06-13/drone-swarms-are-the-new-fireworks-lighting-up-china-s-skies). In response, Intel plans a 1500-drone swarm.

Someone else has unleashed a swarm of 80 boat drones.

(https://i.imgur.com/mmFSZEX.jpg)

I realized we were shitting up the Awesome pictures thread for some reason, so just in case you somehow missed it being posted all over the internet:

Boston Dynamics is teaching their robots how to tea bag your corpse and do a victory dance over it. (https://www.youtube.com/watch?v=kHBcVlqpvZ8)

Also how to chase you up stairs, unlike a Dalek. (https://www.youtube.com/watch?v=LikxFZZO2sk)


Which caused me to catch up on some of their videos.  I actually think this one of spot starting around the 45 second mark is the most creepy:

https://www.youtube.com/watch?time_continue=5&v=tf7IEVTDjng

Everytime I see them do the banana peel trick, this primitive part of my psyche screams out "stop fucking taunting them you fools!"

These things are so cool and so creepy.

I work for a company that makes robots now.

But not dog robots.

Bah. Rags did it better in Woody Allen's Sleeper back in 1973.

There is something particularly disturbing about robot dogs that I am not sure I can put my finger on.  Is it just me?  Robot bear might be worse, but not by much.

Didn't they use robot dogs to chase down dissidents in Fahrenheit 451?

Black Mirror already went  there (https://www.youtube.com/watch?v=xejjA2AFO5I)

I'm almost convinced Boston Dynamics are deliberately copying that look. Fuckers.

And we've been using real dogs to hunt criminals for a while now, and we also have an evolutionary history of being stalked and eaten by wolves.  It'd be reasonable for us to be instinctively wary of things that are dog-sized whose intent we can't read as friendly.

Boston Dynamics has had dog-like robots well before that Black Mirror episode like "Spot" in 2015:

https://www.youtube.com/watch?v=M8YjvHYbZ9w

It will be sad when all the dogs get put out of jobs.

Wiki actually says the Black Mirror guys were inspired by the Boston Dynamic robot dogs to make that episode.

Not this again.  Look, the robot dogs will only come in and do the menial tasks, like fetching the newspaper and scaring off burglars.  You know, the jobs that the regular dogs don't want to do anyway.

If my robot dog will actually just bark at burglars, not piss and shit in the house when it rains and not steal my food when I'm not looking I'll take it.

I've been waiting for someone to invent a robot dog that'll chase feral cats and raccoons out of my yard.

I’m rereading Snow Crash and the Rat Thing starts to read more like science than fiction.

Disney's robot stuntmen (https://www.youtube.com/watch?v=nZ950ywJy0M)

Is it just me or did their demo look a lot like they were just flinging mannequins off a trapeze into a safety net?

Well, mannequins that could strike different poses and then stick the landing perfectly, yeah.

Interesting that the stunt guy interviewed liked the idea, as he didn't see it replacing him but as a versatile tool he could use to set up his stunts more safely.

Good chances that the guy has received many, many blows to the head.  If they can perfect those things (and shit, some of it looks close), there will be much less work for guys like him.  Although, I imagine there are still many kinds of stunts that these robots would not be suitable for.  Maybe even most kinds.

I think what annoyed me about that video was that they were talking about "sticking a landing" while showing video of mannequins flopping ass first into trampolines.  "Sticking a landing" is when you land on your feet without stumbling.
(https://i.imgur.com/t3zXANe.gif)

The midair poses were kinda cool, but you don't really need complex robotics to do that.

Yeah, that was my thought as well.  Not only could they not "stick" the landing, the couldn't land at all.  Still, you can imagine the scenarios where they can use something like this, like anything where they now use cables and pulleys to fling stuntmen (and actors) all about.  Just so long as they don't have to land credibly, I doubt they are even close to figuring that out. 

Pedants, on the internet? What is this?

 :oh_i_see:

Sticking a landing for this means landing on the body (machine) part you planned to land on. And yeah, the robotics looked pretty complex to me, but clearly our expectations are somewhat apart...

I think they're simple relative to Boston Dynamics, but in terms of what Hollywood needs (human-looking body that somersaults realistically and lands in a way that could preserve props and costuming for multiple takes) the results sort of speak for themselves. Apparently Disney's main plan was to have these available for live stunt shows where they'd fall off buildings or get launched through the air in impressive and repeatable ways while keeping their performers safe. That they could also show up in film is simply an added benefit.

In terms of the stuntman, I don't think he's even being naive. Many stunts are relatively minor things like standing somewhere dangerous, running quickly on unsteady ground, falling a relatively short distance. The variety of human actions a conventional stuntman can perform on short notice for relatively low cost is still competitive.

Mostly these acrobatic robots would take over for dangerous stunts which today would likely require CG work, making them slightly more practical and visually realistic.

Don't get me wrong, I think these things are pretty amazing.  But 99 times out of a 100, "sticking a landing" means exactly what the gymnast girl did, so they should not have used it in that context in this article.  I was only making a point that - as impressive as this stuff is -it is a pretty huge gap between what these robots are doing and the incredible balance/muscular control it takes to actually land that sort of movement.  The same can be said about the Evil Robot Dogs.

Stuntpeople generally don't "stick their landings" either. So these robots are no worse in that regard.

Well, not for those big, bombastic jumps they were showing in the video.  But I'd argue that they probably do have to stick their landings for most of the mundane stunts, which is probably the majority of them, wouldn't you think?  I imagine these robots are ridiculously expensive, so in order to be viable they would have to be able to do quite a broad set of activities, one would expect.  I mean, if the only use case is the huge jumps flying through the air and striking a pose, there will much cheaper ways to accomplish that.

(insert joke about how many dead stuntmen you could buy with one robot)

Just being devil's advocate.  Still cool as hell, and any step forward is interesting.

If a stunt performer goes airborne, they land on their back or their ass in something soft, period. They do it for a living and may have to do it several times in a row, you don't get points for staying on your feet.

--Dave

Stuntpeople need to be good at landing in a net on their ass because human spines are vulnerable to damage and landing in a net on your head will break you.  It's a good skill for a human stuntperson to have.

Dummies don't have spines.  If you're throwing a dummy off a building, you don't need to spend money on programming it with net-landing skills because dummies don't have spines.  Most of the time you don't even need to bother with a net.

 ¯\_(ツ)_/¯

I'm sure there's some reason that this is useful for "stunt robots" or they wouldn't have gone to the expense of building it, but the video didn't make it obvious to me.  Seems more like an intermediate "tech demo" step in building something that might someday be useful (hence all the stock footage of human stuntpeople doing cool stunts that the robots are obviously a long way away from being able to do).

The answer may be that there is a lot of money in the entertainment industry.

The robots would not have to be either ridiculously expensive or able to do a broad set of activities. Stunts are dangerous and stunt people are expensive for those reasons. The "mundane" stuff is neither dangerous nor expensive. The ability to have a robot to it saves you significant risk and planning costs for dangerous stunts

Hyundai is building some kind of car robot mech (https://www.cnbc.com/2019/01/08/hyundai-unveils-walking-car-design-that-has-limitless-possibilities.html) for sale.

It's cool I guess, but my gut instinct is that its use cases are so niche that it would be pretty tough to make it commercially viable.  Looks more like something that should be traversing a lunar or martian landscape.

Boston Dynamics parody.  (https://youtu.be/dKjCWfuvYxQ)

Funny, and well made.  I don't know what's more worrying, the fact that we can make such convincing robots, or the fact that we can make such convincing videos.

The guy mo-capping the robot really did a good job of capturing the movement of those Boston Dynamics abominations.

I actually thought most of what I saw was a robot, up until the end anyway.

China launches (http://www.globaltimes.cn/content/1162320.shtml) 50ft "unmanned warship" described as a "mini-destroyer" drone.

Allegedly can conduct air defense, anti-ship and anti-sub missions with phased-array radar, vertical launching system (VLS) and torpedoes.

(https://i.imgur.com/zHh9Nm5.jpg)

Another article (https://www.businessinsider.com.au/china-launches-new-combat-ready-unmanned-warship-2019-8) notes the US military is developing "a high-endurance, reconfigurable ship able to accommodate various payloads for unmanned missions" and has been experimenting with small, crewless "swarm boats".

If you can solve issues around communication and control (aka dealing with jamming/hacking etc) why the fuck would you put squishy humans onboard boats/subs/planes/tanks at all going forward?

This amused me when it was in my timeline yesterday afternoon -

CNET posts a video from Boston Dynamics about 10 of those robot dogs pulling a truck across a parking lot -> tweet (https://twitter.com/CNET/status/1166153356043345922)

So one of the mushers I follow responded with "My dogs can do that!" (https://twitter.com/BlairBraverman/status/1166495709543182336) and it amused me greatly.  Granted, she used her full team of 18 dogs but still... :D

In any conditions that they can endure, I think I'd take a team of sled dogs over robots which just look creepy no matter what.

The robot dogs are not interested in your opinion on the matter.

Happy Judgement Day anniversary, folks.

Long live Skynet, or something.

I, for one, wish our robotic overlords would hurry the fuck up before our human overlords wreck the place.

--Dave

$999 AI drone available to consumers from November: Skydio 2 (https://www.youtube.com/watch?time_continue=48&v=imt2qZ7uw1s). Flies itself and tracks you. "Sees" with six 4K cameras and automatically avoids obstacles.

You basically just tell it to watch you and film you. 3.5km wireless range, 23 minutes' flying time, 36km/h tracking. Can also operate as a regular drone.

Co-worker was just showing me that this morning (we're actually sort of looking into procuring a drone for work use).  Looks really nifty if it works as well as the video claims it does.

In "Oh shit, they're actually coming" news:

https://www.defenseone.com/technology/2019/11/secdef-china-exporting-killer-robots-mideast/161100/ (https://www.defenseone.com/technology/2019/11/secdef-china-exporting-killer-robots-mideast/161100/)

tldr; China selling actual kill-bots. Saudi Arabia and Pakistan are the buyers.

--Dave

Argo AI rolling out robotaxis in Miami and Austin. (https://www.youtube.com/watch?v=3kNPezEsybg)

I would absolutely not get in one of those things at present unless it was on streets with no other cars.

Just watched it California Stop at least twice, which seems like something the wrong cop might throw a fit about.

No Black person to shoot (unless passengers) so probably not.

"It may be a computer, but it's got the confidence of a white person." (https://twitter.com/TheDailyShow/status/1514349633534799872)

That's Cruise in SF. It's been available for the public to ride for a few months now. One of their cars also blocked a fire truck (https://www.forbes.com/sites/bradtempleton/2022/05/27/sf-fire-dept-says-cruise-robotaxi-did-not-yield-properly-and-delayed-fire-truck/?sh=3f0a10c95bb6) responding to call. Waymo also has driverless cars cruising around in SF but those are still being tested and only employees* can ride in them.

* Not sure if that's just Waymo employees or all Alphabet and subsidary employees

All of the driverless cars in Austin are downtown and on the east side around me.

Worst city in america for this. Our roads suck and everything is under construction. Cars are basically handicapped.

People seen this DALL-E (https://openai.com/dall-e-2/) shit?

Since it's not available for the public to play with yet I'm a little skeptical that the proof-of-concept images on that site are generated by the AI without a fair amount of human assistance, but even if it needs handholding, it's pretty cool.

I've heard the images they're touting have been highly curating every step of the process to get the best results, which makes sense.

That said, it's a glorified collage maker as far as I can tell. I'm sure some of these will sell for NFT millions at some point.

Most of the time you'd get something closer to GANfield (https://vdalv.github.io/2018/12/04/ganfield.html). #justMLthings

https://huggingface.co/spaces/dalle-mini/dalle-mini

It's been the entire discord server for two days now. Well this and Midjourney.

(https://i.imgur.com/ymemLPs.jpeg)

I got access to the real DALL-E and fed it the same prompt as a comparison point, lol.

(https://i.imgur.com/PLHCQ3R.png)

This is my favorite thing I've gotten it to make so far: Emperor Norton playing the banjo.  It's like if Matt Berry played him in an episode of Drunk History.

(https://i.imgur.com/ACSFOI3.png)

That's pretty amazing.

Dalle mini fuckin sucks but dalle is better than mid journeys consumer facing shit.

Samwise not being in discord is just disappointment at this point.

Ok zoomer

Anyway, yeah that 'painting' is better than 80% of what I see in exhibits and definitely better than my work  :why_so_serious:

Speaking of robots, apparently Amazon wants to map your homes with iRobots now. People love the surveillance state, yo. If the anti-authoritarians of yore knew how the public was not only willing, but willing to pay for it themselves...

I saw an article about that; fears about it being a play by Amazon to map your home seem a little silly to me since those robots don't have access to any data that your indoor security cameras and/or Alexa haven't already had for years.  Anyone who's bought into the "smart home" thing has already invited so many vampires over the threshold that one more doesn't matter at this point.  And I'm sure there will always be non-cloud-connected alternatives for people who want 90% of the convenience with 0% of the privacy invasion.

Yeah, that already happened with Strava in 2017-2018 (https://www.theregister.com/2018/01/29/strava_heatmap_military_base_locations/) where the 'anonymized' fitness heatmap of the world revealed detailed maps of military bases and troop movements, could be used to identify individual soldiers by name, etc (and also identify where individual Strava users lived). Privacy is such a 20th century concept.  :why_so_serious:

Hell, if someone really wants to map your home using location data, they could use the accelerometers in the phone that you constantly have on your person as you move around your house.  You go places that your Roomba doesn't, and the data would provide insight into how you personally interact with your space, which would be a lot more valuable to advertisers trying to figure out things to sell you for your home than the Roomba-eye-view would.  How much time do you spend in the kitchen in a given day?  How much time do you spend on the couch in front of the TV?  A Roomba can't tell them that, but your phone sure could.  Imagine getting targeted ads for fiber supplements based on the average amount of time you spend on the toilet.

Realistically, though, that kind of thing is very far from being worth the hassle even though it's very close to doable.

I guess we use phones differently.

Most people are using the wireless ones now that you carry around all the time, but if you still have a corded phone that lives on the wall in your kitchen then you're pretty safe.

 :why_so_serious:

I just set the phone down when I get home from work. I don't carry it around, it just sits there until I leave the house. It's a phone.

The phone part of it is probably the least used feature on it for me.  And I think almost everybody else in the world.  So, yeah.....

How I use my phone: Camera. Calendar. Watch. Appointment Book. Address Book. Notebook. Auto Navigation & Trip Planning. GPS. Tracking Beacon. Recipe Book. Music Player. Map of anywhere in the world. Flashlight. Car Radio. Current and forecast Weather for anywhere in the world. Photo Album. Email. Calculator. Alarm Clock. News. Encyclopedia. Timer. NSA Listening Post. Documentation and schematics for pretty much any appliance or gadget I own. Plant Identifier. Spreadsheet. Database. Star Atlas. Simple games for killing time. Bank Teller for most anything not cash. Library. Movie Theater. Shopping. Cat Recording. Bill Paying. Chat & Messaging. Magnifying Glass. Bulletin Board. Exercise/Physical Therapy Coach. Photo Editor. Dictionary. Language Translator. Walkie Talkie. Intercom. Tape Recorder. Menus, ingredients and hours for most restaurants in the country and often place orders for pickup or delivery. Movie and Entertainment Schedules and Ticket Purchasing. Airplane Boarding Pass. Wallet. Scanner. Pedometer. Compass. Receipt Log.

oh, and occasional phone calls.

Yeah, phones are the least necessary except for when they're absolutely necessary. I struggle to think back to the era where you really had to worry about what you'd do if you broke down by the side of the road.

Waymo report after 1 million "rider-only" miles.

Blog post: https://blog.waymo.com/2023/02/first-million-rider-only-miles-how.html

Research paper: https://storage.googleapis.com/waymo-uploads/files/documents/safety/Safety%20Performance%20of%20Waymo%20RO%20at%201M%20miles.pdf

(https://i.imgur.com/Sukiaga.jpg)

"55% of all events were a human driving into a stationary waymo" cracked me up because it's so believable.  I've been in one collision in my entire life so far, and it was some goober backing into me while traffic was stopped.

I'm going to stay wary on this just because there's so much evidence that Tesla hasn't been entirely forthcoming about their own data; there is so much money riding on autonomous driving working well that everybody's got an incentive to shave and manipulate the data, or to control the conditions of testing in ways that are hard to call out without complete access to the testing and development team.

The other reason I'm going to stay wary on this is that I think we're somewhere between 5-15 years before the people selling autonomous driving technology begin to push for it being mandatory in all new cars sold, to be followed by making it illegal to drive manually in many major road networks, on the argument that this will be on net safer. Which it might be, but the values on the other side of the equation won't get any attention at all in that conversation.

I'm sure that info is heavily, heavily slanted/cherry-picked i.e. incidents happening when a Waymo vehicle was stationary would still cover where one had stopped unexpectedly or parked itself in an unusual place. Likewise I could see them classifying a mototrist taking evasive action as 'behaving dangerously' or violating highway rules. There have been a number of reports of Tesla's issues with their self-driving experiments and, frankly, these results sound so much in advance of all the issues most self-driving efforts have that it just sounds pretty fishy.

What I really don't understand (profit) is why all these efforts aren't going in to producing an automated public transport system with a wider selection of routes rather than an inefficient and pollution wise far worse option of a system of private taxis. I personally don't have an issue with having an automated driving solution if I was confident the tech was up to the challenge but I'm not sure we're going to be getting to that point in the next few years.

With Tesla, it's easy to see why: Musk has a profound ideological hatred for public transport. With the other people in the market, yeah, I don't quite see it.

There's a book called Aramis, Or the Love of Technology, by the recently deceased French intellectual Bruno Latour that is about a really interesting (if expensive and before its time) French prototype system that was basically an autonomous-driving approach to light rail--basically you'd build cheap light rail infrastructure all through the suburbs and you'd have little self-driving cars on rails that you could order via a telephone/early Internet system and as they moved towards denser areas with more connections, they'd link up and move on to the central spinal spokes of the system and then kick off to a higher speed for faster connection between suburbs and the center city. It failed because the tech just wasn't there but right now I think something like it is plausible--and you wouldn't need full autonomous AI able to adapt to every condition on our existing road infrastructure.

Tesla deliberately hobbled themselves early in the game by deciding they were going to do everything with visual processing alone (which I think was an Elon edict rather than a decision made by somebody with any technical knowledge), and on top of that way over-promised their capabilities (again, Elon).  They don't belong in any conversation about self-driving tech at this point except to say "here's everything not to do".


I'm not sure if you actually needed an answer, but yeah, profit.  We don't invest in public transportation in this country, not in any serious way, and building a whole new system like what you're talking about would be a very serious investment.  If you put a lot of engineering effort into building an automated public transport system (limited utility anyway, because public transportation is already optimized to move a lot of people with a minimal number of operators), nobody's going to buy it.  If you can sell a luxury car that doesn't need a chauffeur, or run a taxi service that doesn't need to pay drivers, you will make billions, if not trillions.

(edit) It turns out Waymo published a detailed research paper (https://storage.googleapis.com/waymo-uploads/files/documents/safety/Safety%20Performance%20of%20Waymo%20RO%20at%201M%20miles.pdf) that analyzed and described all of the incidents in their million miles of autonomous driving.  It'd be nice if they included the video footage for each incident as well so you could judge for yourself whether their reporting is accurate, but I'd imagine there are privacy/liability concerns involved with publicly shaming bad drivers, and I would also imagine that if they'd mischaracterized or omitted any severe incidents, there would be publicly available police reports and/or lawsuits to prove it.  So it seems reasonably safe to trust the facts that they're reporting, even if skeptical of any spin they might try to put on them.

The most severe collision involved the Waymo being stopped at a red light (which is pretty reasonable, right?) and being rear-ended at 25mph by a driver who was looking at their cell phone.

The second most severe collision was one where a slower-moving vehicle cut the Waymo off and immediately braked; the Waymo did the right thing and applied its brakes, but obviously if a stationary obstacle suddenly appears in front of you there's only so much the laws of physics allow.

Based on the raw numbers, and without even getting into at-fault details, it seems like Waymo is about twice as safe as the average human driver (2 crashes in a million miles, compared to about 4 for the average human).

The target market mindset for all of these self-driving things is clearly the rich asshole who disdains public transit because of having to mix with all the poors, and not the city planner who would love a more cost effective and efficient way to move citizens and relieve his traffic problems.

That goes back decades to the entire way we incentivize how our cities and towns were built. The cult of the automobile and the individual freedom it represents has absolutely destroyed any real commitment to public transit. We don't build for it, we build to put fucktons of cars on the road driving into urban centers only when they need to for work and never stopping until they return to their suburbs. With those parameters, self-driving cars for a passenger capacity of one totally makes sense as something to do.

To paraphrase that Blizzard dildo, "Don't you guys have cars?"

Income inequality in the US is higher now than it was at the peak of the Roaring Twenties a hundred years ago, which makes rich assholes the most lucrative target market they have ever been.

Not saying it's a good state of affairs, but nobody should be asking "why would you make this instead of designing a public transit system that nobody will ever fund".  That question should just answer itself.

The NYT has a piece today on how every single consumer products manufacturer and retailer is abandoning the mass market as a source of profitability--everybody is chasing "premiumization" as the main way to improve the profit margin, e.g., making the main product as shitty as possible and making anything even slightly resembling quality only on an upmarket basis.

It's an insane strategy in that wealth inequality is even starting to erode the upper middle-class; a billionaire's family can only consume so much caviar or caviar-priced goods. But yeah, this is the world that America is charging towards. So even though in a better timeline, there could be trillions to be made on building a better public transport system with autonomous piloting/driving, that would take valuing the greatest good for the greatest number and clawing back wealth from the .01%.

My wife was telling me recently about a vendor sales pitch that was talking about a feature that they didn't have yet but it's "on their roadmap for next quarter," and how she couldn't believe that some of the suits in the room were falling for that, because obviously "on the roadmap" doesn't guarantee it'll be there in any particular timeframe.

"Not only that," I said, "but if they finally add the feature, who says they won't put it in a more expensive SKU than whatever they initially sold you?" and she was like "holy shit you're right, that's exactly what they'll do."  I hate that I've worked in tech long enough to be able to see these things coming a mile away, but the entire economy is built around that kind of bullshit now.

Let's not forget how terrible people are about owning cars. Treat them like shit, drive them like shit, maintain them like...I don't want to insult shit here...And then get a new one as soon as they've paid off the old one, or earlier.

I'm trying to baby my poor old truck into her senior years, avoiding salty slush, regular maintenance (I wish I had a good classic car garage in the area to work on her, though). And when I look at folks who make almost double what I do (not difficult tbf), but have dropped well north of a hundred grand on vehicles with no end in sight to the treadmill. And then complain they are broke all the time (and in layers of debt). Ironically many of the folks I'm referring to also drive domestics that need a ton of repairs over their life (I finally got new brakes a couple years ago...truck's 15 now).

The car economy is crazy toxic and nobody seems to give a shit. Or at least not enough to change their behavior, whether it's driving better, maintaining the vehicle, or owning it until it actually needs replacement.

All that said, in my area, taking public transportation is a truly shitty experience. The drivers are terrible, the routes make no sense, and the past few years have made dealing with the average person such an ordeal that I'm happy most of my public transportation experience were in the sublime metros of Central LA and South Oakland. Not kidding, 'the poors' on those buses and trains were way more well-behaved.

I'd still do public transportation over a robot car, though. I don't even like the level of electronics in my truck, and it's completely without screens.

I do all my own car maintenance that I can handle without investing $$$ in specialty tools. Even then, vehicles these days are way more complicated than the stuff I grew up tinkering with in terms of standard tune-up oil change type stuff. I take care of the filters and oil changes and other minor items that might go haywire, but some of this stuff you are forced to take to a dealer/mechanic. Of course there is youtube for damn near everything, but that means investing time.

So I can excuse folks for not keeping up with maintenance if they have to take it somewhere that just end up being an argument about the upcharges.

Shit, I have to argue for them to complete the maintenance I ask for, every time. In all my years of faithfully maintaining the truck, I've had one instance of a place upcharging on wipers, which NYS law kinda sets up for them with a 6-month replacement requirement. It must be the new thing, same happened to the old lady (this was just last year). But that was for the NYS inspection, not a maintenance visit. I go by the maintenance schedule in the book, which gives both time (every 6 months) and mileage. My main issue after getting a place to actually perform all the maintenance listed, necessary or not, is getting them to obey the time of ownership metric vs the mileage metric.

The burden of the low mileage vehicle owner  :grin: (I should be rolling over 64k on the '08 soon, though)

it's not just cars though.  Everything is disposable these days.  Clothing, furniture, houses, appliances, food.  Built cheap from the cheapest possible materials.  And there are very very few alternatives available for most things because they just don't sell. At least 9 out of 10 people will pick the item on the shelf with the lowest price without even looking at how well it is made or of what materials.  It is an odd paradox. Our culture practically defines itself by what we own, yet we buy junk.

People buy junk because that's all there is now. You could have a million dollars to drop on a toaster but the only way you're going to find a well-functioning durable toaster is if you use that million to get an engineer to build you a great one from scratch.

Most durable goods are now assemblies of the same cheap parts and components with some kind of outside shell that gives them some degree of brand distinction, but is otherwise the same as any other version. They're all made to break fairly quickly. No brand is made from high-quality parts that the brand-name company assesses for quality under its own control except maybe cars, in fact. I was talking with a guy who maintains a bunch of rental properties for relatively demanding higher-end clients and he said that they literally can't find a dishwasher model for the properties that is well-made and durable, so they've just settled for a basic cheap model and they throw them away the first time they break because fixing or maintaining them isn't worth it.

That's true now, but that's because for the last 40 years Americans have been buying the cheapest thing on the shelf even though it was only going to last 5 years rather than paying 15-20% more for the well-built one next to it that would last the rest of their lives.  Since well-built stuff doesn't sell, nobody makes it anymore.  There are a few brands out there that are still relatively well-made - Bosch appliances for example.  But now the price differential is 200-300% instead of 15-20%.

It's not Americans, this is the world.

For Bosch, you have to buy the high end of their product line to get an appliance that's mostly assembled in Germany with higher quality parts. (You also have to avoid their power tools line, which sucks.) You buy lower end Bosch at Home Depot or whatever, you're getting the same as all the other brands. Miele is still pretty much "this is really well built" and that too you pay through the nose for. But that's big durables. You cannot get well-built smaller appliances any longer for love or money. Nobody makes them, period.

It's almost as if the amount of people who can afford the priciest shit isn't enough to sustain a business on, especially when you are building products that aren't designed to be re-bought in a few years' time.

Well, shareholder dividends can't keep going upupUpUPUP if the consumer isn't buying new stuff every few years.  It's almost as if consumer capitalism is dedicated to eating itself.

Global capitalism isn't even shareholder capitalism any more. The financial world is effectively run by three hedge funds now, who no longer remotely think about their investments in terms of company performance or fundamentals or anything of the sort. Mostly they're just the financialized upper tier of the global economy with private equity being the middle tier, and private equity is mostly about looking for companies to buy and liquidate.

I got full access to Waymo a couple of weeks ago, and took my first ride in one today (fully unmanned; it pulled up in front of my house and dropped me off at a restaurant).  Observations in no particular order:

• Its route avoided not only freeways but also major streets, I'm guessing to minimize the odds of encountering something unpredictable, since any accident is a major liability right now regardless of fault.
• As far as navigating the route it chose, it was flawless.  Smoothly avoided double-parked cars, slowed down perfectly as it came to speedbumps, and very gracefully handled a two-way stop crossing a busy multi-lane street that would have stressed me out as a driver.  The only time it did something that felt "off" to me was lurching slightly while stopped at a red light, and I realized it did that to try to dodge a motorcycle that was coming too close for comfort, moving out of the way of the reckless human as far as it could without leaving the lane or crossing the limit line.  Asimov's First Law in effect.
• My prior hands-on experience with an AV was a Tesla with autopilot, which was a white-knuckle ride the entire time that it was on.  The difference isn't day and night, it's day and the deadly vacuum of space.
• The display inside the car shows you everything the car is "seeing," including pedestrians (which show up as dots) and their dogs (which show up as smaller dots) as well as static obstacles like curbs and medians.  A fun detail is that anything which it identifies as an obstacle that requires extra attention (like a pedestrian that's hovering too close to the curb as if they might dash out into traffic) is highlighted in green.
• I was riding with my visiting mother-in-law, who to give you an idea of her demographic has never used Uber, and she couldn't stop talking about how cool it was.

I doubt I'll be using it regularly, for all the same reasons I don't use Uber regularly, but as a tech demo it was absolutely wild.

That's cool.

I am still thinking the short-to-medium term scenario is going to be all the money that's deep into autonomous driving is going to insist on (and pay for, grudgingly) the installation of guide technology for autonomous vehicles along major routes that will especially include HOV lanes on major highways; step 2 in about 15 years will be to argue that manual vehicles should have to get out of autonomous-access lanes for safety; step 3 will be making highways autonomous only except in low-density areas, at which time trucking is going to be shifted to autonomous only for highway transport to big depots where human drivers might do pick up for local delivery. (Might).

There's persistent evidence that Tesla in particular is concealing a lot of data about accidents and erratic behavior; I think the other players are avoiding some of that by grooming the routes they're testing with undisclosed guided trials. E.g., the routes they perform on well are a kind of "Clever Hans" thing where they've been trained quietly a zillion times on about the aberrations and unusual features of the route, they're not genuinely adaptive AI except when it comes to handling unexpected behavior by other vehicles/pedestrians, etc. I think the fix is going to be to exert control over the conditions of the test, so to speak.

They're never going to get human-controlled vehicles off of small rural routes or highways in severely underpopulated states within the next 50 years, but highly trafficked highways? Human drivers are going to be moved off whether the tech works as planned or not.

It's hard to imagine "unusual features" of a route that would be harder to figure out adaptively than what people and vehicles can do, and it's equally hard to imagine that people are out there manually coding in the locations of each pothole and speedbump when LIDAR is able to detect things like that, but with enough VC money and stupidity anything is possible.

Tesla is in a class of its own because Elon decided LIDAR would detract from the clean aesthetic lines of the Cybertruck and so everything has to be done with cameras -- the results speak for themselves, I think.

Or, we could just lay down these nice pieces of metal, maybe make them parallel to each other. String some wires about 20 feet above them and then put a series of wheeled tubes between said wires and metal.

Shared infrastructure?  Sounds like socialism.

(edit) tbh even in a car-free utopian future with ubiquitous rail, I can see municipal robotaxis being a very solid last-mile solution for those who aren't physically able to ride jerk scooters and the like.

Buses and trains will never work, no. Only hyperlooped AI tunneling machines.

I'm glad I'm not the only one smart enough to think of things like this.

https://abc7news.com/amp/contra-costa-county-micro-transit-system-glydcars-autonomous-vehicles/14125023/


I've thankfully never lived as far out as any of those places, but am familiar with the commute situation from talking to coworkers and it's grim -- if you want to take BART, you get up early to secure a parking space at the station before they all fill up.  Eliminating parking as the bottleneck in getting people off the freeways and onto the rails is pretty huge.

Hands on with Gemini (https://youtu.be/UIZAiXYceBI?si=QO4upgsJr1AuKO2h)

I'm very skeptical of the claims about Gemini's "reasoning" abilities until the general public gets to poke at it, but the visual recognition stuff alone is pretty impressive.

Like you say, I want to see what happens when people are fucking with it. Or just doing stuff that isn't planned. "Our favorite interactions" is kind of like "selected game footage"--until you see the rest, you can't be sure how much it resembles the typical experience.

This is faked. And not just in the disclaimer that's not in the video itself but only in the description "For the purposes of this demo, latency has been reduced and Gemini outputs have been shortened for brevity." faked.

According to Bloomberg https://archive.ph/ABhZi (https://archive.ph/ABhZi)

I mean, that's... whatever.  If the only gap between what's in the video and reality is speech recognition (which, you know, exists) and throwing more CPUs at it to make it faster, then it's effectively there.  I'm more talking about whether these prompts they're giving it are scenarios that it's been specifically trained on.  A general purpose AI that can play the shell game based on reading about it on Wikipedia is one thing; an "expert" that has been specifically taught to play that game is impressive (object tracking is hard!) but it's in an entirely different ballpark.

Company loses millions after employee duped by video call deepfakes. (https://www.huffpost.com/entry/hong-kong-deepfake-meeting-scam_n_65c27630e4b069b665dd0aa8)

Somebody tell this guy that this whole AI fad is going to blow over any day now.   :why_so_serious:

Wow. I noted the deepfake Biden robocalls (https://mashable.com/article/biden-ai-deepfake-robocall-life-corporation) the other day. A full video meeting scripted setup is next level... but totally doable.

How about a Let's Play series that fakes that last four presidents?  https://www.youtube.com/@GamerPresidentsUSA

Pod Save America uses deepfake Biden for comic relief, but uncomfortably https://www.youtube.com/watch?v=RyqtXPkM364