Slashdot Mirror


Video Games Are So Realistic That They Can Teach AI What the World Looks Like (vice.com)

Jordan Pearson, reporting for Motherboard:Thanks to the modern gaming industry, we can now spend our evenings wandering around photorealistic game worlds, like the post-apocalyptic Boston of Fallout 4 or Grand Theft Auto V's Los Santos, instead of doing things like "seeing people" and "engaging in human interaction of any kind." Games these days are so realistic, in fact, that artificial intelligence researchers are using them to teach computers how to recognize objects in real life. Not only that, but commercial video games could kick artificial intelligence research into high gear by dramatically lessening the time and money required to train AI. "If you go back to the original Doom, the walls all look exactly the same and it's very easy to predict what a wall looks like, given that data," said Mark Schmidt, a computer science professor at the University of British Columbia (UBC). "But if you go into the real world, where every wall looks different, it might not work anymore." Schmidt works with machine learning, a technique that allows computers to "train" on a large set of labelled data -- photographs of streets, for example -- so that when let loose in the real world, they can recognize, or "predict," what they're looking at. Schmidt and Alireza Shafaei, a PhD student at UBC, recently studied Grand Theft Auto V and found that self-learning software trained on images from the game performed just as well, and in some cases even better, than software trained on real photos from publicly available datasets.

87 comments

  1. let's play Global Thermonuclear War! by Joe_Dragon · · Score: 2

    What side do you want?

    China
    France
    Russia
    UK
    USA
    India
    Israel
    Pakistan
    North Korea

    1. Re:let's play Global Thermonuclear War! by Moheeheeko · · Score: 1

      France, just because "Force de Frappe" is fun to say

    2. Re:let's play Global Thermonuclear War! by Anonymous Coward · · Score: 0

      Canada, eh? Fear our maple syrup warheads!

    3. Re:let's play Global Thermonuclear War! by PopeRatzo · · Score: 1

      Canada, eh? Fear our maple syrup warheads!

      True story: the Canadians named their intercontinental ballistic missiles after Guy Lefleur.

      [OK, it's not really a true story, but it would be so cool if it were.]

      --
      You are welcome on my lawn.
    4. Re:let's play Global Thermonuclear War! by Anonymous Coward · · Score: 0

      It's Guy Lafleur. And he's french Canadian so his name isn't pronounced "Guy" as in "that guy over there". FIY.

    5. Re:let's play Global Thermonuclear War! by Anonymous Coward · · Score: 0

      The only winning move is not to play?

  2. All good then by Anonymous Coward · · Score: 0

    As long as the world conforms to what artists think it should look like and the sun cooperates by lighting it correctly, I'll have no fear of driving into wall and other such mild nuisances.

    1. Re: All good then by WarJolt · · Score: 1

      Google should train an AI to drive using GTA and stick it in one of their cars. It will drastically reduce the time it will take to train autonomous vehicles.

    2. Re: All good then by Anonymous Coward · · Score: 0

      and kill more prostitutes

    3. Re: All good then by St.Creed · · Score: 1

      They are already planning for that. The codename for the AI has just been published, it's "Christine".

      Sounds really nice!

      --
      Therefore, by the (faulty) logic you're using, you're just a cow with a keyboard - osu-neko (2604)
  3. Does the AI know fear... by __aaclcg7560 · · Score: 1

    I wonder what the AI would think of Doom as the monsters hide behind secret wall panels and jump out from behind to attack?

    1. Re:Does the AI know fear... by 110010001000 · · Score: 2

      AI doesn't think at all. What we call "AI" is really just a bunch of algorithms.

    2. Re:Does the AI know fear... by Anonymous Coward · · Score: 0

      AI doesn't think at all. What we call "AI" is really just a bunch of algorithms.

      You don't think at all. What you call "you" is really just a bunch of algorithms.

      No, I don't actually think that AI today is able to "think" in the same way that humans think. I just hate the "it's just a bunch of algorithms" argument, because you cannot prove that you (or I, or any human) isn't just following "a bunch of algorithms".

    3. Re:Does the AI know fear... by dmbasso · · Score: 0

      You don't think at all, you're just a bunch of neural circuits.

      --
      `echo $[0x853204FA81]|tr 0-9 ionbsdeaml`@gmail.com
    4. Re:Does the AI know fear... by 110010001000 · · Score: 1, Troll

      So human intelligence is just a bunch of algorithms? When did YOU get your Nobel?

    5. Re:Does the AI know fear... by K.+S.+Kyosuke · · Score: 0

      You'll never convince such religious nuts. They'll always resort to special pleading. Just because.

      --
      Ezekiel 23:20
    6. Re:Does the AI know fear... by 110010001000 · · Score: 1, Insightful

      Religious nuts? That is what you AI and Space Nutters are. You think that somehow AI is going to magically appear because your PC got a lot faster between 1990 and 2016.

    7. Re:Does the AI know fear... by dmbasso · · Score: 1

      Ah, that explains it... I haven't realized that was the origin of his statements, but I guess you're right. The irony in his "Oh so you know how the brain works?" is that I've actually been studying it for a long time, so yeah, I have a good idea about the generals. :p

      But after the "Wow, when did you receive your Nobel?", it's easy to see how his brain works - it doesn't! :p

      --
      `echo $[0x853204FA81]|tr 0-9 ionbsdeaml`@gmail.com
    8. Re:Does the AI know fear... by K.+S.+Kyosuke · · Score: 0

      Confusing people again?

      --
      Ezekiel 23:20
    9. Re:Does the AI know fear... by 110010001000 · · Score: 1

      If you think the brain is just a bunch of "neural circuits" then you are stuck in 1960. We tried neural nets back then. Didn't work.

    10. Re:Does the AI know fear... by 110010001000 · · Score: 1

      Sorry you are confused. At least you got your dose of anger out for the day :-)

    11. Re:Does the AI know fear... by amicusNYCL · · Score: 1

      You really bought in hard to that whole "things will never change" mantra, didn't you? Is there some bookie that will let you bet your 401k on progress not occurring, or something? Why are you so insistent on trying to explain to people how their fantasies of progress occurring are based on mythology and magic?

      Congratulations on your Nobel, by the way.

      --
      "Our two-party system is like a bowl of shit looking at itself in a mirror." - Lewis Black
    12. Re:Does the AI know fear... by K.+S.+Kyosuke · · Score: 1

      I'm not the one complaining about "AI and Space Nutters" in every other comment. BTW, shouldn't you be frantically writing to news outlets right now?

      --
      Ezekiel 23:20
    13. Re:Does the AI know fear... by ArylAkamov · · Score: 1

      Exactly what an A.I. would say, Mr. 110010001000. I'm watching you.

    14. Re:Does the AI know fear... by dmbasso · · Score: 3, Insightful

      We tried neural nets back then. Didn't work.

      It seems it is you that are stuck in 1960, because connectionist techniques nowadays are nothing like that. And deep-learning has been breaking record after record, even achieving superhuman performance in some tasks. And deep-learning is just a smart math trick over the regular backprop algorithm, allowing more layers without degrading the error gradients. When the models that actually incorporate neuroscientific knowledge (current research) mature, expect even better performances.

      And wtf, this "I tried once, I failed, I'll never try again" is surely a loser-talk.

      --
      `echo $[0x853204FA81]|tr 0-9 ionbsdeaml`@gmail.com
    15. Re:Does the AI know fear... by amicusNYCL · · Score: 1

      Wait, you're not trying to suggest that we didn't know all that there was to know in 1960, are you? We were using chemical propellant for rockets then, are you really trying to suggest that we did NOT understand all that there was to understand back in 1960? That maybe we were missing something? Because that doesn't sound like you at all.

      No, you're wrong about this. Chemical propellant is the pinnacle of space launch technology, the brain is a bunch of neural circuits, we already knew all of this in the 60s, it's time to accept this universal truth and get on with our lives.

      --
      "Our two-party system is like a bowl of shit looking at itself in a mirror." - Lewis Black
    16. Re:Does the AI know fear... by Anonymous Coward · · Score: 0

      Well since APK has retired as the most annoying poster on Slashdot 110010001000 has gleefully stepped into his shoes, and is even surpassing his irrelevant and unrelated interruptions into threads.

      I suspect he was savagely attacked by an Astronaut and Deep Blue at some point and is holding a grudge.

    17. Re:Does the AI know fear... by 110010001000 · · Score: 1

      You aren't complaining because YOU are a Space and AI nutter! It is like a homoerotic thing with you :-)

    18. Re:Does the AI know fear... by 110010001000 · · Score: 1

      Things will change. Just not the things the Space and AI nutters are counting on. Welcome to Earth.

    19. Re:Does the AI know fear... by 110010001000 · · Score: 1

      "Deep learning" is the 2016 version of "Neural Nets" to get some more VC money. Meanwhile it is some voice interface hooked up to Wikipedia.

    20. Re:Does the AI know fear... by 110010001000 · · Score: 1

      The brain is not a bunch of neural circuits. I already said that. You are a moron. And it is very possible that chemical rockets ARE the only launch technology possible. If there was something better we would be using it rather than using chemical rockets FOR THE LAST 60 YEARS. You guys are incredible. You seem to think there is going to be some magical breakthrough that is going to launch you to Mars.

    21. Re:Does the AI know fear... by Anonymous Coward · · Score: 0

      I don't know about the systems you deal with but no system I've used has gotten faster in any way that is directly beneficial to me. My computer has 'twice' the memory, 'twice' the speed, 'twice' the storage. My computer is asked to process 'four times' the information I could before before I am actually allowed to see content.

      Sure, I have 'twice' the memory, adobe acrobat reader used to be 5 MB. Now it is 128 or so MB just for the installer. My and 99% of people's use for acrobat reader is static .pdf viewer. 99% of the security problems with acrobat reader are because of crap 99% of the user base doesn't need. Don't get me started on programs that specifically require acrobat as the default pdf reader rather than having A default pdf reader.

      Extra power means nothing if the person using the tool doesn't get anything for it. If all the extra power is going to process targeted marketing, artificial 'ad auctions' before page display, downloading audio/video with auto play enabled, then its just an arms race. Folks are starting to check out of that shit in certain ways.

    22. Re:Does the AI know fear... by Atzanteol · · Score: 1

      What's the difference?

      --
      "Ignorance more frequently begets confidence than does knowledge"

      - Charles Darwin
    23. Re:Does the AI know fear... by Anonymous Coward · · Score: 0

      AI doesn't think at all. What we call "AI" is really just a bunch of algorithms.

      You don't think at all. What you call "you" is really just a bunch of algorithms.

      No, I don't actually think that AI today is able to "think" in the same way that humans think. I just hate the "it's just a bunch of algorithms" argument, because you cannot prove that you (or I, or any human) isn't just following "a bunch of algorithms".

      cogito ergo sum. knowing I think is the only thing I know for sure everything else is based on inference deduction and making guesses about what is casting those shadows on the wall of Platos cave. My guesses are probably accurate but I cannot "prove" them. It is just simpler to assume that you are a thinking being akin to myself than a very advanced eliza bot. Unfortunately I cant prove I think to you so lets just agree we are to scripts pointlessly replying to eachother

    24. Re: Does the AI know fear... by Anonymous Coward · · Score: 0

      People going "ah-oom ah-oom", and getting razor sliced, locked up, beaten up and killed for being " different" in 1960

    25. Re:Does the AI know fear... by goose-incarnated · · Score: 1

      Wait, you're not trying to suggest that we didn't know all that there was to know in 1960, are you? We were using chemical propellant for rockets then,

      I'm curious - what are we using now as rocket propellant?

      As far as the AI argument goes: The processing power in the world the last 3 decades has increased by a factor of a few hundreds of thousands. The "AI" has improved by a factor of 2. Maybe less. The progress we've seen in AI software has been incremental; a rounding error compared to progress in other software or in hardware, and we've yet to approach the cognitive abilities of a smart cockroach, even though we're consuming orders of magnitude more processing power than the cockroach.

      Luckily, we can continue brute-forcing our way through a problem-space, pruning the space here and there, and get pretty good results because of all the transistors devoted to the task. The singularity isn't coming along in my lifetime (nor yours, in case you're a young man).

      --
      I'm a minority race. Save your vitriol for white people.
    26. Re:Does the AI know fear... by amicusNYCL · · Score: 1

      Oh, OK. So an "AI nutter" apparently believes that AI will "magically appear", because PCs got faster over 26 years. I didn't realize that you could distill their beliefs down to something that sounds so stupid, but I guess you're the expert.

      --
      "Our two-party system is like a bowl of shit looking at itself in a mirror." - Lewis Black
    27. Re:Does the AI know fear... by ChrisMaple · · Score: 1

      If it walks like a duck and quacks like a duck...

      --
      Contribute to civilization: ari.aynrand.org/donate
    28. Re:Does the AI know fear... by ChrisMaple · · Score: 1

      ln -s /usr/bin/xpdf /usr/bin/acrobat

      --
      Contribute to civilization: ari.aynrand.org/donate
    29. Re:Does the AI know fear... by amicusNYCL · · Score: 1

      If there was something better we would be using it rather than using chemical rockets FOR THE LAST 60 YEARS.

      Listen, when we were doing that thing where we parodied each others' positions in the other story, and I said exactly what you just said, that was supposed to be comedy, man. That's satire. The notion that we would have known in the 60s all of the physics necessary to accomplish any kind of travel through space/time is laughable, ridiculous, and obviously satire. You shouldn't repeat it like you actually think that, it makes you sound incredibly short-sighted. You are turning into that person where it's not possible to tell if it is satire or someone being serious.

      And it is very possible that chemical rockets ARE the only launch technology possible.

      OK, for the sake of argument let's take something that you were using as satire: rail guns. What part of a miles-long electromagnetic track sloping up is physically impossible according to the laws of physics? As far as I can tell it's prohibitively expensive (given our current budget priorities), and will come with a massive set of additional and unique engineering challenges, but I'm missing the part where it's violating laws of physics that would make it not even possible, and chemical rockets still the only way. So, what is it about an electromagnetic launch system which makes it impossible? Are you suggesting that since we didn't do that in the 60s that it is therefore not possible to do at all, ever? Because I can't begin to try to explain the number of things wrong with an assertion like that, things that are trivially and obviously disproven just by observing things around you, and if you actually believe that shit then I don't know why you bother posting at all. In fact, you should really go whole-hog and only use communication technologies that were available in the 1960s. It's like you're trying to argue that it's not possible to use a "sky crane" with retro-rockets to gently and autonomously lower a vehicle weighing several tons on the surface of Mars and then have it drive around for several years taking pictures and other measurements and sending data back to a relay in orbit before transmitting on to Earth, because we didn't do that in the 60s. I'm not sure what your reason is for believing that if things didn't happen in the 60s then they won't happen ever, but it's a weird belief to have and to defend so rabidly.

      You seem to think there is going to be some magical breakthrough that is going to launch you to Mars.

      You're the only person using the word "magic". What does that say?

      The brain is not a bunch of neural circuits. I already said that. You are a moron.

      Uh huh. So if you believe that the state of neuroscience in the 60s is the final stage, then you're a moron, but if you believe that the state of astrophysics in the 60s is the final stage then you're some enlightened being. You need to be more consistent in your belief system before you try to get others to join your cult, or else they're going to think you're just a troll.

      --
      "Our two-party system is like a bowl of shit looking at itself in a mirror." - Lewis Black
    30. Re:Does the AI know fear... by K.+S.+Kyosuke · · Score: 1

      Your "logic" clearly says that anyone interested about the future of the human species in any way must either carefully cleanse his mind of all AI or spacefaring progress thoughts or he'll be a space and/or AI nutter. Well, guess what, this is your brand of "homoeroticism", whatever that means, not many other people's.

      --
      Ezekiel 23:20
  4. Re:Not the same by Anonymous Coward · · Score: 1

    If you think GTA is photorealistic, you are, quiet simply, a fucking idiot ;)

    Says the guy who doesn't know the difference between quite and quiet. Not to mention that a winking emoji is not a proper way to close a sentence. Just to be clear: you're missing a period there, Mr. not-a-fucking-idiot.

  5. Skynet Loves Ragdoll Physics by Anonymous Coward · · Score: 0

    Skynet will get such a huge kick out of ragdoll physics that it won't be able to try it out on real humans.

  6. Is this how crappy AI is? by Anonymous Coward · · Score: 0

    You can plug it into a video game and it gets better at predicting images?

    There is a problem with our scientists today. The SR-71 is still the fastest plane ever built. Saturn V is our best rocket. Concorde is our best plane.

    Feynman, Einstein and Newton would have intellectually destroyed the people we have today who are posing as scientists.

    1. Re:Is this how crappy AI is? by 110010001000 · · Score: 2

      Although you are right about the poor quality of what passes for "research" nowadays, people need to realize that there are limits to technology. It doesn't increase in power and get better and better indefinitely. There are limits based on Physics and engineering. Particularly with digital computers we are hitting those limits now. CPUs in particular are only getting marginally faster with each (expensive) generation.

    2. Re:Is this how crappy AI is? by Anonymous Coward · · Score: 1

      It isn't the hardware that's the problem - it's the software and the Wetware

      Faster processors can help a little, but better algorithms are what is required. Apparently there is no money in AI. The best minds of our generation are being paid to work out how to get a user to click on an advert.
       

    3. Re:Is this how crappy AI is? by amicusNYCL · · Score: 2

      The SR-71 is still the fastest plane ever built.

      That doesn't mean that it's not possible to build a faster plane. There's no reason to build one at this point. You don't need a plane capable of outrunning any missile when you can take pictures with a satellite instead. You don't even need a plane that fast for carrying people when we haven't even attempted to handle the legislation involved with supersonic passenger travel. We didn't lose any capability after the SR-71 was retired, the capability had already been replaced by the time the last ones were laid up.

      And, there most definitely are faster aircraft than the SR-71. The SR-71 is the fastest manned air-breathing aircraft. There are faster manned aircraft that are not air-breathing (mach 6.7 X-15), and there are faster air-breathing aircraft that are not manned (mach 9.6 X-43).

      The SR-71 met a specific goal at a certain time and performed exceedingly well at that mission, but we haven't replaced it with a better plane because we don't need one. That's like saying that naval scientists or shipbuilders have lost their art because the last battleship was launched in 1944. We just don't need them anymore.

      --
      "Our two-party system is like a bowl of shit looking at itself in a mirror." - Lewis Black
    4. Re:Is this how crappy AI is? by Kjella · · Score: 1

      You don't even need a plane that fast for carrying people when we haven't even attempted to handle the legislation involved with supersonic passenger travel.

      Uh, Concorde? From what I understood it's the market that killed it, with modern communication via email and video conferencing the business market became considerably smaller. You of course have the luxury market but they don't necessarily travel between major business hubs very often, they're often going more exotic places. It needs economic solutions to bring the cost down and technological solutions for the sonic boom so it can go over land and not be such a special case, the legislation is the least problem.

      --
      Live today, because you never know what tomorrow brings
    5. Re:Is this how crappy AI is? by Anonymous Coward · · Score: 0

      The best minds of our generation are being paid to work out how to get a user to click on an advert.

      Most accurate assessment award goes to.... AC.

      Mockery, Sarcasm and Irony are all-in one for the win so why didn't you logon to comment AC?

      Sigh...

    6. Re:Is this how crappy AI is? by Anonymous Coward · · Score: 0

      We don't even *build* horse-drawn carriages anymore!

    7. Re:Is this how crappy AI is? by AK+Marc · · Score: 1

      There is a problem with our scientists today.

      The problem is capitalism. Optimizing for cost gets a technically inferior solution.

      The SR-71 is still the fastest plane ever built.

      It's the fastest manned aircraft that launches from the ground under its own power and runs on JP-7 fuel. Remove any one of those constraints, and there is something faster, though none are in use, as they are too expensive, and anything you can do with an SR-71, you can do cheaper with a drone or satellite. The SR-71 existed so that we could take pictures without losing pilots. Drones fit that bill and are much cheaper.

      Concorde is our best plane.

      Grounded because they were expensive and delicate. They were inferior to a 777 in every way, other than travel time.

    8. Re: Is this how crappy AI is? by Anonymous Coward · · Score: 0

      (Not parent) The great part about slashdot is I've been on this site for 10+ years, participated in many conversations, and NEVER needed to make an account. I refuse to create one now because I would be down voted just for having too many digits in my account ID.

    9. Re:Is this how crappy AI is? by drinkypoo · · Score: 1

      Although you are right about the poor quality of what passes for "research" nowadays, people need to realize that there are limits to technology. It doesn't increase in power and get better and better indefinitely. There are limits based on Physics and engineering. Particularly with digital computers we are hitting those limits now. CPUs in particular are only getting marginally faster with each (expensive) generation.

      The brain is a parallel system. Computers are getting more parallel. There's a long way to go before we throw up our hands and say "this is impossible". If you were running the show we'd never have got out of the caves.

      --
      "You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
    10. Re:Is this how crappy AI is? by amicusNYCL · · Score: 1

      The legislation is a major man-made problem. If you have planes flying all over the country producing sonic booms, people are going to sue and that's what is going to stop it. Your suggestion of a technological solution to stop the sonic boom sounds great, but I don't see how that's possible without engines producing virtually no sound yet still strong enough to propel something several hundred tons faster than the speed of sound.

      --
      "Our two-party system is like a bowl of shit looking at itself in a mirror." - Lewis Black
  7. Recognizing objects by 110010001000 · · Score: 0, Troll

    AI "researchers" have been trying to train computers to "recognize objects" for the last 50 years. They seriously need to just give it up. AI isn't happening. The horsepower isn't going to be there with digital computers. Digital computers are hitting a dead end with the end of Moores Law. We won't have enough power to do the computations required using the current "AI" methods. Time to try something new.

    1. Re:Recognizing objects by Anonymous Coward · · Score: 0

      God you're an idiot. But you already knew that - you insist on getting people to point it out to you on almost every article when you post your willfully ignorant trollish shitposts.

      That said... this article is just as stupid as you are. They've been doing this for decades. In fact, that's why we HAVE a 3d game industry.

    2. Re:Recognizing objects by 110010001000 · · Score: 0

      Oh we HAVE a 3d game industry because...it can teach AI? Amazing!

    3. Re:Recognizing objects by Anonymous Coward · · Score: 0

      AI "researchers" have been trying to train computers to "recognize objects" for the last 50 years. They seriously need to just give it up. AI isn't happening.

      No, it'll happen eventually. It's just that you're too stupid to imagine it.

      Just like all the other amazing things that were said to be impossible or impractical, including flight to the Moon or the portable microcomputer you use to jack off to horse porn every day.

      AI will happen, maybe not in the next few months or years of your shortsighted stupid-o-vision, but it will happen.

    4. Re:Recognizing objects by Anonymous Coward · · Score: 0

      So you know with certainty the extent that computing can reach? When did you get your Nobel prize?

    5. Re:Recognizing objects by AK+Marc · · Score: 1

      Where is the "dead end" with moore's law? When will we hit it? We have enough compute power for AI. What we don't have is a good algorithm.

  8. Closed by OpenSourced · · Score: 1

    So, on one side, you have the game program converting conceptual objects from a database to 3D images, using a powerful GPU in the process.

    And in the other side you have the AI program taking that 3D image and converting it to a conceptual object, and putting it in a database, using a powerful GPU in the process.

    And then you wonder about global warming.

    --
    Rome taught me patience and assiduous application to detail. Virtues which temper the boldness of great, general views.
    1. Re:Closed by UnknownSoldier · · Score: 1

      You're comparing apples and oranges.

      One is a forward transformation.
      The other is the reverse transformation.

      Mapping between the two is non-trivial.

    2. Re:Closed by K.+S.+Kyosuke · · Score: 2

      Why is this so strange to you? Virtual worlds have been used for simulating inputs into control systems and the like since at least the times of Apollo (feeding simulated data into the AGC back then).

      --
      Ezekiel 23:20
    3. Re:Closed by 110010001000 · · Score: 0

      You missed the point. As usual. I warned him no one would understand what he was talking about.

    4. Re:Closed by K.+S.+Kyosuke · · Score: 2

      No, both of you missed the point. Simulating stuff like this is cheaper, faster, simpler, therefore more efficient than working with the physical world. It is the physical alternative that would cause more global warming than this.

      --
      Ezekiel 23:20
    5. Re:Closed by K.+S.+Kyosuke · · Score: 1

      Every activity has costs. Things cost time and money. Time means people people need to commute, get fed, etc. Food and fuel production and consumption contributes to global warming. Likewise, other expenses like materials, manufacturing, etc. cost money. All that money likewise boils down to energy and labor costs, and the latter again is ultimately an energy expenditure. So it turns out that any costs are roughly proportional to energy expenditures. So limiting your costs by getting things done faster and cheaper means they have proportionally smaller environmental footprint, not larger.

      --
      Ezekiel 23:20
    6. Re:Closed by 110010001000 · · Score: 1

      I think you missed the point. Read it again. He isn't saying simulation is bad - he is saying THIS method is bad. Seriously you need to learn to comprehend what other people are saying.

    7. Re:Closed by AK+Marc · · Score: 1

      No, we all understood, and recognized it as a red herring. You aren't testing what happens inside the conceptual environment. Doing the double-transformation would be silly to run a simulation to "train" the computer in basic operations. The double-transformation is done to train the recognition of the intermediate step of "seeing" a wall and recognizing it as such. The AI isn't learning how to interact with the world, that's pre-programmed, and doesn't need the intermediate step.

      The intermediate step is to train the AI to recognize a world-recognizable wall. That can't be done without the double-transformation (at least until it's well trained, then that can be rolled up in the initial knowledge).

    8. Re:Closed by drinkypoo · · Score: 1

      I think you missed the point. Read it again. He isn't saying simulation is bad - he is saying THIS method is bad.

      He's still wrong. Here's the original comment:

      So, on one side, you have the game program converting conceptual objects from a database to 3D images, using a powerful GPU in the process.

      And in the other side you have the AI program taking that 3D image and converting it to a conceptual object, and putting it in a database, using a powerful GPU in the process.

      And then you wonder about global warming.

      The rebuttal is not that it doesn't consume energy, but that this is a brain damaged way to look at it because no comparison is given. It's just a complaint that energy is being used, but it doesn't come with any consideration of what might be more efficient.

      On one hand you've got the cost of putting the AI into a thing and then having that thing rove around the real world, and the cost of shepherding it around to specific situations, and/or creating specific situations for it to study, and on the other you've got the cost of spitting out a simulation for it. I know from experience that you can do that with less than 500W sustained. How much energy will it cost just to deliver your AI-bot to its destination[s]? Answer, more.

      Seriously, you need to learn to comprehend the argument.

      --
      "You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
    9. Re:Closed by ChrisMaple · · Score: 1

      Current technology of CMOS gates shows roughly the same energy per transition as a neuron (given similar levels of complexity.) Since AI uses so much power for such marginal results, this implies that computer hardware is not properly designed/optimized for intelligence work or that AI software is woefully wrong (or both).

      Brain cells and brains as a whole aren't magical; they work by some mechanism. Equivalents of all mechanisms can be made by digital logic systems, but we don't know how to make the equivalent of a human brain / nervous system yet. When we do, we should expect equivalent performance.

      A number of barriers exist.

      • We don't really know what's going on yet.
      • Brains are much more parallel than transistor systems, and replacing parallel with multiple serial is energy inefficient.
      • Brains are densely packed in 3D; transistors are densely packed in 2D and the third dimension by comparison is 100 to 1000 times larger (A neuron is about 1 micron. IC wafers are about 1 mm thick, although they can be shaved down.)
      • At least some long-term learning in brains occurs by rewiring, rewiring in digital circuits is unusual (reprogrammable FPGAs) and inefficient - digital circuits learn by storing data.

      Not only is there a long way to go, it seems to me that a GPU is spectacularly inappropriate for AI work

      --
      Contribute to civilization: ari.aynrand.org/donate
  9. GTA by SeaFox · · Score: 2

    Good. We need AI that can correctly identify a person as a prostitute and various fictional weaponry and automobiles.

    1. Re:GTA by Nidi62 · · Score: 1

      Thank god they aren't using Saints Row or we'll have AI robots running around using dildo bats.

      --
      The only thing necessary for evil to triumph is for it to be pitted against a slightly greater evil
    2. Re:GTA by Anonymous Coward · · Score: 0

      That's a bad thing?

  10. Re:Not the same by Anonymous Coward · · Score: 0

    It's bitztream, the autism-hating Slashdot troll!

  11. Re: Not the same by Anonymous Coward · · Score: 0

    How sad is your life that you feel better about pointing out a typo of two transposed letters and emoticon

  12. Re:Not the same by Anonymous Coward · · Score: 0

    If you think GTA is photorealistic, you are, quiet simply, a fucking idiot ;)

    If you think that visual sensors used for AI training are photorealistic, so are you.

  13. Re: Not the same by amicusNYCL · · Score: 2

    An interrogatory sentence should end with a question mark.

    --
    "Our two-party system is like a bowl of shit looking at itself in a mirror." - Lewis Black
  14. Fantastic! by WolfgangVL · · Score: 2

    Now the fleets of self driving cars will be sipping "hot coffee" and beating the hookers to collect their money back for services rendered.
    They will also know how to rack up a respectable 5 start wanted level and just hide under a bridge till it all goes away.
    Oh, and my favorite new autonomous feature? Spawning attack choppers out of thin air to aid in robbing every convenience store in the city while performing the above two tasks.

    --
    You are being ripped off every second of every day, so that advertisers can help rip you off even more tomorrow.
    1. Re:Fantastic! by Anonymous Coward · · Score: 0

      Man, self-driving cars sure sound fun.
      I can't wait for the future. I've been waiting all my life for it!

  15. Re:Not the same by AK+Marc · · Score: 1

    Realistic enough to work as a proxy for reality, but not realistic enough to fool an alert observer. Apparently you don't know the difference between "works" and "perfect". There's usually quite a gap.

  16. wow by Anonymous Coward · · Score: 0

    it sure is fun listening to people who don't know how AI works projecting their human, real-world based interpretations onto them. Never do see that in the mainstream media.

  17. What will you do when AI tells you... by Anonymous Coward · · Score: 0

    ... that white people have the right to have their own countries? And that you're all brainwashed idiots, who are actually incapable of the most BASIC thought processes?

    Please explain why you think white people, and ONLY white people, should accept millions of non-whites into our countries every year, thus turning our countries into multi-racial hellholes for white people. (Apart from "The TV told me I'd be a 'bad person' if I spoke out against mass immigration"...)

  18. Minecraft as a training world? by Anonymous Coward · · Score: 0

    Makes sense, the world is big, made to be explored, a dare I say, photorealistic?

  19. Talos Principle! by JThundley · · Score: 1

    The Talos Principle is coming to life!