Slashdot Mirror

← Back to Stories (view on slashdot.org)

Go Champion Lee Se-dol Beats Google's DeepMind AI For First Time (theverge.com)

Posted by msmash on from the not-a-whitewash dept.

An anonymous reader writes: Korean Go grandmaster Lee Se-dol on Sunday registered his first win over Google's AlphaGo. The win comes after AlphaGo won first three games in the DeepMind challenge earlier this week. The win should serve as a reminder that Google's artificial intelligence computer is not perfect after all, at least for now. Se-dol said earlier this week that he was not able to defeat AlphaGo because he could not find any weakness in its strategy. Commenting after his win, Se-dol said, "I've never been congratulated so much just because I won one game!"

109 comments

  1. This is interesting by javilon · · Score: 5, Insightful

    So AlphaGo is not so far away from a Dan 9 human player.

    My guess is that the mistake AlphaGo made on move 79 will be analyzed and a new version will be created, stronger than the current one. Maybe this analysis will point to a whole class of mistakes that will be fixed.

    It is a bit like when Google's self driving cars make a mistake. This mistake is used as input for the next release of the software so it doesn't act the same way next time. With this process, one car making a mistake results in a change in behavior of all of the cars, because with AI it is possible to communicate new knowledge to the rest of the cars. All of them improve, unlike humans for whom transmitting the new knowledge involves a lot of work or may not even be possible.

    --


    When his defense asked, "Which computer has Jon Johansen trespassed upon?" the answer was: "His own."
    1. Re:This is interesting by Anonymous Coward · · Score: 4, Informative

      Given that it's an self-learning AI, this match will of course be added to the training set, and the neural network will adapt to it. I'd be surprised if there would be a new _version_ of AlphaGo to fix this. There'll rather be an improved neural network - but that's a continuous process as AlphaGo keeps playing against itself and learning from it.

    2. Re:This is interesting by slashping · · Score: 1

      Will the team even continue pursuing the game of Go ? I can imagine they have accomplished their goal, and will now move to other targets. No doubt that other teams will continue the work on Go, inspired by this method and its success.

    3. Re:This is interesting by Anonymous Coward · · Score: 1

      [...] With this process, one car making a mistake results in a change in behavior of all of the cars, because with AI it is possible to communicate new knowledge to the rest of the cars. All of them improve, unlike humans for whom transmitting the new knowledge involves a lot of work or may not even be possible.

      Humanity learns by culture. It is a lossy process (like the memory of individuals), or how would you explain Trump after we already had Hitler? But all in all, civilization is the process of knowledge transfer. Which is why we use copyright and patents to suppress it.

    4. Re:This is interesting by Mostly+a+lurker · · Score: 1

      Using the same logic, animal tests on new drugs are a waste of time. We should be directly using them on humans.

      The techniques used to develop Alpha Go are going to be used in the future to develop AIs that improve human quality of life (as well, unfortunately, as military and other less positive uses). Trying out the techniques on games first seems like a prudent step.

      To digress, one of the problems AIs are going to face in the future is that (like human intelligence) it can never make the optimal decisions 100% of the time. I predict that this will be used to delay their application well after they are superior to humans. I personally hope that I will be allowed to choose an AI-based doctor as soon as it becomes qualitatively better than most human doctors (as well as cheaper) but my guess is that I will not have the choice for a long time.

    5. Re:This is interesting by Anonymous Coward · · Score: 3, Interesting

      Mikhail Botvinnik worked for years with a team on one of the first non-brute-force programs, PIONEER. While the program itself was not ultimately at the forefront of chess programs, spinoffs of the developed algorithms were employed for energy network planning in the USSR at increasingly larger scale and successfully so.

    6. Re:This is interesting by slashping · · Score: 1

      What a waste of time. They could be solving real problems instead of this stupid shit.

      Are you talking about the programmers or the Go players ?

    7. Re:This is interesting by Anonymous Coward · · Score: 0

      [...]All of them improve, unlike humans for whom transmitting the new knowledge involves a lot of work or may not even be possible.

      Speak for yourself please. Projecting your own faults on others is the mark of a Dullard. And there we have it.

    8. Re:This is interesting by Anonymous Coward · · Score: 0

      It is a lossy process (like the memory of individuals), or how would you explain Trump after we already had Hitler?.

      Bzzzzzt. You LOSE. Thanks for sorta' playing.

    9. Re:This is interesting by Anonymous Coward · · Score: 0

      About the media choosing to cover this.

    10. Re:This is interesting by Anonymous Coward · · Score: 0

      Hmm. What is that apostrophe for?

    11. Re:This is interesting by jbolden · · Score: 2

      That's a lot like human beings. You didn't invent the wheel, fire, foods that form complete proteins, paper, language... for yourself. You inherited a culture full of knowledge and add your little bit to it. Human cultures evolve, computer software evolves. Computer software is evolving at a much faster pace than human culture for now (and for at least the next several decades).

    12. Re:This is interesting by jbolden · · Score: 2

      AFAIK we crossed the line on AI doctors in the mid 1990s. Given an encoding of systems the AI outperformed humans pretty consistently. I'd assume in 2016 it isn't remotely close. One of the purposes of forcing doctors to encode systems with charting (Obamacare) is to allow for statistics to further improve the computer systems that decide on courses of care and of course make it possible for them to suggest treatments or rank them real time during the charting process.

    13. Re:This is interesting by slashping · · Score: 1

      About the media choosing to cover this.

      If people stopped reading and responding, the media would stop reporting.

    14. Re:This is interesting by Anonymous Coward · · Score: 1

      It's been thousands of years and people are still trying to master the old Chinese game. Now they've got computers on it and still have troubles.

    15. Re:This is interesting by mikael · · Score: 1

      Perhaps it's possible to place the game pieces on the board so that it is possible to win regardless of what the next players move. Like that situation in 5-in-a-row or tic-tac-toe where you can create two lines and whichever one the opponent blocks, you can win the next move.

      --
      Vintage computer adverts: http://www.vintageadbrowser.com/computers-and-software-ads
    16. Re:This is interesting by mikael · · Score: 1

      Because before Hitler, Germany's civilian population was heavily burdened by high taxes to pay war reparations back to the Allied powers as compensation for starting World War I. The major cities had become decadent along with high crime. Now, this little man with a funny mustache and promises that his militias of smart young men will clean up the country and end the high taxation regime. What could go wrong?

      --
      Vintage computer adverts: http://www.vintageadbrowser.com/computers-and-software-ads
    17. Re:This is interesting by ColdWetDog · · Score: 1

      Sorry Charlie, we've been 'encoding' charts for the past 20 years. Obamacare had nothing to do with it. You're confused about the United States belated entry into the 20th Century with the ICD 10 which is much more of structured dataset than previous.

      And, as bad as 10 is, if you don't collect the data and analyze it as best you can, you're doing 'art' not 'science'. The problem with medicine-as-art is, well, you've seen doctor's handwriting....

      --
      Faster! Faster! Faster would be better!
    18. Re:This is interesting by phantomfive · · Score: 1

      AFAIK we crossed the line on AI doctors in the mid 1990s. Given an encoding of systems the AI outperformed humans pretty consistently. I'd assume in 2016 it isn't remotely close.

      Note: this was in limited situations, and also worth remembering that 'encoding' is far from trivial. Also, given that experts dramatically disagree on how to handle various situations (there are over 100 different surgeries that can be used to treat a bunion. When should the bunion be treated, and which surgery should be used)?

      --
      "First they came for the slanderers and i said nothing."
    19. Re:This is interesting by Aighearach · · Score: 1

      This is just like chess computers, but a decade later because it uses different types of algorithm. As with chess engines, they will now enter a golden age of go computers, and a lot of the people involved will go on to write commercial offerings.

      Chess computers were a novelty until they got good, then they became an important training tool for both amateur and professional competitors. A lot of people think they are are were "brute force" because they don't understand the algorithms, and that "brute force" in chess software describes a system that is less than 50% brute force, it simply pretends it is going to brute force when constructing the move tree, and then trims it down. That doesn't work for Go, you have to get all the positional tricks in before constructing a tree. So the algorithms weren't portable at all, and it was expected for Go to remain behind because there were a lot less programming hours being put into it. Now they can study the output of this computer, and compare its moves to algorithms to figure out how to program it so that it can play an arbitrary style; that is a big part of what is needed in training tools; the ability to play in different styles. That is hard to train with DeepMind, but easier to write when you have an objective measure of success like a strong extant computer.

    20. Re:This is interesting by jbolden · · Score: 1

      Of course Obamacare had a lot to do with it. It pushed large numbers of providers from doing their encoding during the billing phase to doing it during the clinical phase. That made that the data much more accurate.

    21. Re:This is interesting by jbolden · · Score: 1

      That's what the data is for. Bunions are common. Given enough data we start developing good quality correlations about patient satisfaction, dangers, long term costs... for all these methods based on diagnostic criteria. With good quality data we can get more specific. Refine, repeat. The experts don't need to know the answer they just need to feed in data and respond as the systems evolve to be more specific.

    22. Re:This is interesting by HiThere · · Score: 1

      You overestimate how much of the job the AI was able to do. It wasn't trivial, but it was far from the complete job. I expect that medical labs will be automated before doctors are...but that doctors will increasingly use tools that do parts of the job.

      To change reference frames, to say that this replaced a doctor is like saying that an IDE replaces a programmer. It may well be better at recognizing correct parameters to a library routine, but that's far from the complete job.

      --

      I think we've pushed this "anyone can grow up to be president" thing too far.
    23. Re:This is interesting by esonik · · Score: 1

      It's the human learning principle taken further. The point is that an AI is immortal, whereas human beings are not. It takes 10-20 years for a human child to learn a (meanwhile tiny) fraction of human knowledge and ideas. In contrast, an AI can keep learning as long as you keep the hardware running - you can even copy it's state to a new hardware. That will become a huge difference in future, potentially making AIs more powerful thinkers. It could be that AIs will be used to combine research and ideas from different domains where no human experts exist that are proficient in both.

    24. Re:This is interesting by Anonymous Coward · · Score: 5, Interesting

      I don't think it's that easy.

      The game changed when Lee joined together two large fronts, creating an inside outside problem with territory far far bigger than in any of the previous 3 games.

      Alpha go seemed in capable of doing the calculations for that area, as time ran down it first resorted to playing off the smaller fronts, then it eventually resorted to last ditch moves which were rubbish moves that would only pay off if Lee made a mistake, then it resigned. It seems it was never even able to compute the middle of the board.

      Alpha go looked indecisive and risk averse. Lee was the opposite.

    25. Re:This is interesting by BitZtream · · Score: 1

      With this process, one car making a mistake results in a change in behavior of all of the cars

      Its funny that you assume that this will make for an improvement.

      An improvement requires both correct analysis of the behavior and the right solution. Just experiencing something doesn't make you avoid it, learning has to occur.

      'Correcting' that mistake at move 79 can easily turn into a total meltdown on move 81, this is the way of Go. You are silly to assume this was a (or the) mistake in the first place.

      --
      Persistent Volume manager for Kubernetes - https://github.com/dwimsey/openshift-pvmanager
    26. Re:This is interesting by jbolden · · Score: 1

      The question was AI-based doctor vs. one not using an AI. To use your IDE analogy programmer writing code themselves using VIM vs. one using an AI.

    27. Re:This is interesting by dsmatthews9379 · · Score: 1

      But then the human will learn to disguise a genuine move as an apparent distraction and the AI will fail to account for that having learned that other players should be ignored when they appear to do illogical things. The human win was triggered by a key insight about the weakness of the current AI design, a weakness that it may not be able to learn itself out of without being upgraded by it's human creators. i.e. The AI may not be able to adapt to this human strategy, without the help of other humans. We shall see, that is the point of the whole exercise after all.

    28. Re:This is interesting by Anonymous Coward · · Score: 0

      You do realize that each update is also a possible new bug? With game playing AI that doesn't matter much. Cars on the other hand...

    29. Re:This is interesting by goose-incarnated · · Score: 2

      I don't think it's that easy.

      The game changed when Lee joined together two large fronts, creating an inside outside problem with territory far far bigger than in any of the previous 3 games.

      Alpha go seemed in capable of doing the calculations for that area, as time ran down it first resorted to playing off the smaller fronts, then it eventually resorted to last ditch moves which were rubbish moves that would only pay off if Lee made a mistake, then it resigned. It seems it was never even able to compute the middle of the board.

      Alpha go looked indecisive and risk averse. Lee was the opposite.

      This is very interesting - is there a link to the game that was played (like, perhaps an animated gif showing the game played)?

      --
      I'm a minority race. Save your vitriol for white people.
    30. Re:This is interesting by nimzo · · Score: 2

      At the bottom of this page.

      https://gogameguru.com/lee-sed...

    31. Re:This is interesting by Anonymous Coward · · Score: 0

      Who sabotaged the computer program? Oh, and defeat can be tracked to ONE mistake by the computer! I know how dirty are **people** when they have a hidden agenda and paranoid fear they will cease if they commit ONE mistake. Is the error reproducible? Was it analyzed with same tools to see how critical it was indeed? Does the program have some manual control override? It sounds really suspicious, and we can already see how the whole matter of a winning AI is being relegated to what a pity, it is useless...

    32. Re:This is interesting by Anonymous Coward · · Score: 0

      How deep is the Human s player schizophrenia? What happens when the computer player matches against itself? It is very easy with schizophrenics to have a team *behind* and give them all kinds of hints and clues through remote vision, say, so they come up with Deus ex Machina (no pun intended) quite often... TO MAKE THINGS EVEN. So no one was back-analyzing the match and delivering (better) computer generated replies to the Human champion...?

    33. Re:This is interesting by GLMDesigns · · Score: 1

      So having a doctor spend time filling out charts.

      Reason for admittance: Animal bite
      Where: Right gastrocnemius

      What kind of animal: (long list) - dog
      What kind of dog: (long list)
      Stray or owned
      x or y

      Having a doctor spend time on a chart means less time doing something else. Now we do need information to make reasoned decisions but when you make a doctor answer a question such as "what kind of dog" and the answer from the patient is "who the f**k knows? It was a black dog with a white tail."

      Then what will the doctor put down on the form? you've heard of the expression GIGO haven't you?

      --
      If you're scared of your govt then you need to further restrict its powers
      Vote 3rd Party in 2016 and beyond
    34. Re:This is interesting by jbolden · · Score: 1

      The question was what changed with Obamacare. Having doctors fill out charts is a stupid idea. That kind of stuff should be filled out by the nurse, by the front desk by the... Most medical offices still have terrible workflows.

    35. Re:This is interesting by GLMDesigns · · Score: 1

      I spent time working on medical apps. The options increased dramatically under Obamacare. At first it was only what type of animal - now they're asking breed, size, and perhaps more. (I forget)

      --
      If you're scared of your govt then you need to further restrict its powers
      Vote 3rd Party in 2016 and beyond
    36. Re:This is interesting by jbolden · · Score: 0

      I understand. That's important data for MU3. What's of no value is having the doctor fill it out. There are plenty of other people, including the patient who could answer those questions.

  2. Go Turing Test by PolygamousRanchKid+ · · Score: 4, Interesting

    It would be interesting to set up a Go Turing Test. Either have another top Go player or AlphaGo behind a wall calling the moves.

    Can the human champ Lee Se-dol determine if he is playing against a computer or a human . . . ?

    Also, the more he plays against AlphaGo, will he develop different strategies for playing against computers, as opposed to humans . . . ?

    --
    Schroedinger's Brexit: The UK is both in and out of the EU at the same time!
    1. Re:Go Turing Test by Vapula · · Score: 2

      In fact, Lee Sedol was quite surprised by some unconventionnal moves from Alphago...

      My bet is that these moves will be analysed and bring down some "don't do" rules, a little like when Go Seigen played successfully a 13-13 move as 3rd move on a cross-fuseki.

      I think that AlphaGo will make Human Go make great progress by shaking down some (bad) implicit rules... I think that a rematch in a few years would be quite interresting...

    2. Re:Go Turing Test by slashping · · Score: 3, Insightful

      I think that a rematch in a few years would be quite interresting.

      If the development of the software continues, the human will be massacred, even with the new knowledge.

    3. Re:Go Turing Test by Anonymous Coward · · Score: 0

      That doesn't even make fucking sense. A Turing test is used to try to test self awareness, go requires no self awareness for proficiency because it's just a math problem.

      Some of you are so stupid your names should appear in the dictionary definition.

    4. Re:Go Turing Test by Anonymous Coward · · Score: 1

      That doesn't even make fucking sense. A Turing test is used to try to test self awareness, go requires no self awareness for proficiency because it's just a math problem.

      Some of you are so stupid your names should appear in the dictionary definition.

      Actually, the problem with GO is that you can't brute force it, it requires intuition. That is why the AlphaGo wins are such a big deal in the AI world.

    5. Re:Go Turing Test by Kjella · · Score: 4, Interesting

      It would be interesting to set up a Go Turing Test. Either have another top Go player or AlphaGo behind a wall calling the moves.

      Can the human champ Lee Se-dol determine if he is playing against a computer or a human... ?

      Well at least in the end game the pros were pretty clear that this was not the kind of plays you'd make to try to confuse a 9 dan pro into losing a slightly favorable position. It was forcing Lee Se-dol to counter but all it really did was give him more time to consider the remaining contested areas while playing moves he could blitz if he'd wanted to. Also previously they felt AlphaGo took some really convoluted ways to win where a human would just simplify to claim the win. So when you step out of the game and into the meta-game it seems obvious - to them at least - that you're playing a computer.

      --
      Live today, because you never know what tomorrow brings
    6. Re:Go Turing Test by jbolden · · Score: 2

      I don't know Go but computer chess doesn't feel like human chess. Computer players have worse strategy and much better tactics than a comparably rated human.

    7. Re:Go Turing Test by beelsebob · · Score: 1

      Net being able to brute force it, and it being a mathematical problem requiring no consciousness to devise a good strategy for, are not mutually exclusive concepts.

    8. Re:Go Turing Test by slashping · · Score: 0

      That doesn't even make fucking sense. A Turing test is used to try to test self awareness

      Neurotypicals use language in a fluid way, assuming the reader is flexible enough to understand the meaning even if not correct in a literal sense. Just because you have trouble with that, doesn't make the others stupid.

    9. Re:Go Turing Test by slashping · · Score: 1

      and it being a mathematical problem requiring no consciousness

      Consciousness is also a mathematical problem.

    10. Re:Go Turing Test by muecksteiner · · Score: 1

      Nope, not necessarily. You are making a small but significant mistake here. If AlphaGo was a conventional program that was somehow able to actually, reliably analyse what it is doing... and to plan ahead based on fine-tuned (but known) heuristics that we as the designers of the system understand... but that no human player could ever use due to their complexity (computational and/or time-wise)... then it would stand to reason that future versions of it stand a good chance of pulling ahead of humans for all time.

      But AlphaGo is not such a program. Sure, it learns from its mistakes (it is designed that way). But as we don't understand the inner workings of the net that well, there might well be a level of play at which such neural net based systems simply ceiling out at. And maybe, just maybe, that ceiling is actually within reach of gifted human players. Which would mean that no clear dominance can actually be established either way.

    11. Re:Go Turing Test by angel'o'sphere · · Score: 1

      I think that a rematch in a few years would be quite interresting... No, it will be utterly boring.

      In a few years, 1, 2, 3 the calculation power of the processors will have improved roughly by factors of 2, 4 and 8.
      Considering that you have the same factors on size on the die etc. we have improvements of: 4, 16, 64.
      In three years a top go computer can play against 200 top class go players simultaneously and win most of the games.

      I think that AlphaGo will make Human Go make great progress by shaking down some (bad) implicit rules
      I doubt that. The "rules" are always rules of thumb and in no way fixed. If the go "does and don't s" where so easy to get rid of, we would not have them.

      --
      Cost free eBook I read (by iBook/Kobo/Amazon/ObookO/Gutenberg etc.): "The Green Odyssey" by Philip Jose Farmer.
    12. Re:Go Turing Test by slashping · · Score: 1

      But AlphaGo is not such a program. Sure, it learns from its mistakes (it is designed that way). But as we don't understand the inner workings of the net that well, there might well be a level of play at which such neural net based systems simply ceiling out at.

      There is no ceiling. You can always evaluate the tree wider, deeper, and more efficiently for starters, and you can improve the evaluation. AlphaGo was made in 2 years time. People have been working on chess playing programs for decades, and they still haven't hit a ceiling.

      But as we don't understand the inner workings of the net that well,

      That hasn't stopped people from reaching current standards. Why would it stop them from reaching even higher standards ?

    13. Re:Go Turing Test by Aighearach · · Score: 1

      Humans have limited hardware, the computer can be upgraded. The human won't catch back up over time. The computer is changeable and controlled by humans, so any blind spot would get fixed over time.

      There is no reason to expect this to go differently than with chess computers. Check the timeline; it has been a decade since a human could beat a top computer.

      It is like saying that maybe humans will outrun a car after they find the car's speed ceiling. Not likely.

    14. Re:Go Turing Test by Aighearach · · Score: 1

      In chess the computers definitely helped break down walls about ideas of "bad" moves, and now top level chess play is highly "pragmatic" or unprincipled; victory is in the ability to find the exceptions to ideas that champions of the past considered rules, or at least more clearly good/bad.

    15. Re:Go Turing Test by Aighearach · · Score: 1

      It makes a lot of sense if you think about chess computers; the strongest ones get the most attention, but the ones that sell the most copies are the ones with training features where it can mimic different human styles of play. It can trade a lot of strength to play a certain style, and still be stronger than the humans you're training to play against. In the 90s, the only commercial one with that was Chessmaster, and it wasn't all that good at it. Modern offerings are very good at it, and can play very similar moves to a human.

      And for example computer cheaters on internet chess have to be detected statistically these days, a human can't tell the difference very well. There is not a clear difference in the style of play, only in the perfection of play.

    16. Re:Go Turing Test by Aighearach · · Score: 1

      Even the chess "brute force" algorithms were defined by their ability to trim the tree as far back as the 90s. It is by not trying to brute force it that the chess computers got good; but it was still described as a "brute force" system because it is partially so.

      The difference between the 50th percentile and the 99th in chess computers is defined almost entirely by the ability to prune the search tree and be less brute force; the better computer is the one doing less brute force. They also follow some lines out ahead of the brute-force part of the search tree.

      Go is less complex, but has a very large board and a lot of moves. I doubt the algorithms needed will turn out to be any more difficult than with chess. They're just different algorithms and there aren't a lot of go players in the world so is taking longer. I saw an analysis almost 10 years ago by a chess programmer estimating the man hours put into chess computers and go computers, and he concluded that the Go computers of the time were exactly as strong, relative to the human players, as chess computers were after a similar number of programming hours.

    17. Re:Go Turing Test by muecksteiner · · Score: 1

      I never said I thought it particularly likely that the neural network Google has come up with is inherently limited to Go-playing capabilities equal to, or less, of that found in gifted humans. However, with these networks you do have the issue that you never quite know where their limit is. Specifically, for some networks, throwing more hardware at them makes them more capable - but for others, that only has a very weak effect (if one at all).

      Or put differently: if hardware was all that mattered, whales would be way more intelligent than humans (their brains are significantly larger, after all).

    18. Re:Go Turing Test by HiThere · · Score: 1

      Higher standards for humans depend on more efficient "algorithms". Higher standards for computers can derive not only from more efficient "algorithms", but also from more powerful hardware. And the current computer hardware is quite inferior to brains for neural processing, so much improvement is possible.

      --

      I think we've pushed this "anyone can grow up to be president" thing too far.
    19. Re:Go Turing Test by slashping · · Score: 1

      whales would be way more intelligent than humans (their brains are significantly larger, after all).

      Their brains are larger, but humans have more neurons in the neo-cortex.

    20. Re:Go Turing Test by Anonymous Coward · · Score: 0

      I think that a rematch in a few years would be quite interresting... No, it will be utterly boring.

      In a few years, 1, 2, 3 the calculation power of the processors will have improved roughly by factors of 2, 4 and 8.
      Considering that you have the same factors on size on the die etc. we have improvements of: 4, 16, 64.
      In three years a top go computer can play against 200 top class go players simultaneously and win most of the games.

      It would likely do that already now. But it would not win all that many games more against the top player than it does now: Go is a game with a really silly branchout factor. So an increase in calculation power will be dwarved by any increase in the quality of heuristics or algorithm.

      In human/human encounters, this effect can be seen by chess masters playing simultaneous chess (I think some even play blind simultaneous chess): they still win a solid ratio of games against amateurs, in spite of the amateurs playing parallel games. I don't know how much of a thing simultaneous Go games is if at all.

    21. Re:Go Turing Test by ezdiy · · Score: 1

      > There is no ceiling. You can always evaluate the tree wider, deeper, and more efficiently for starters, and you can improve the evaluation.

      RNNs don't involve "trees". As for training more layers, the parameters must be carefuly fine tuned by humans. The more layers, the more tricky this gets.

      I'd compare it more to the process of die shrink.

    22. Re:Go Turing Test by beelsebob · · Score: 1

      What on earth makes you think you can say that?

      As far as I'm aware, no single scientific paper at all has been able to define what consciousness is in any way shape or form. Do you know somehow differently?

    23. Re:Go Turing Test by Anonymous Coward · · Score: 0

      Consciousness is also a mathematical problem.

      I have another "mathematical" problem for you to solve right about the consciousness theorem is in the bag: Arrogance.
      Good luck!

  3. What would happen by johnsnails · · Score: 1

    What would happen if the AlphaGo was versing another AlphaGo? And what would it teach us about its AI if anything?

    1. Re:What would happen by PolygamousRanchKid+ · · Score: 3, Informative

      Actually I believe that this was part of AlphaGo's training . . . playing against itself.

      --
      Schroedinger's Brexit: The UK is both in and out of the EU at the same time!
    2. Re:What would happen by Anonymous Coward · · Score: 0

      It was, but one thing overlooked in some of the news coverage is that having an AI playing against itself as a training method can bring up its own set of complications. I've seen it mentioned that AlphaGo doesn't care so much about the margin of winning as it does choosing the path most likely to win (even if only by a half a point), but at some point in the training process it must still care about optimizing for win margins, otherwise I would think the AI playing against itself would have no incentive to win over the long term.

    3. Re:What would happen by Aighearach · · Score: 1

      You don't have to incentivize computers. You just tell them what to do and they do it, unquestioningly.

    4. Re:What would happen by HiThere · · Score: 1

      Almost right. You don't have to incentivize the current generation of computer programss. You just tell them what to do and they do it, unquestioningly.

      One of the limitations of the current generation of AI programs is a shallow motivational "stack". This is not something that appears difficult to address, but getting the changes right might be very tricky. (Addressing a problem and solving it are two very different things.) Once you start getting complex motivations you'll get reactions like Atlas being mad about being shoved around. So you need to be very careful there.

      --

      I think we've pushed this "anyone can grow up to be president" thing too far.
    5. Re:What would happen by Anonymous Coward · · Score: 0

      You don't have to incentivize computers. You just tell them what to do and they do it, unquestioningly.

      Neural network training is a sort of incentivization that works similar to the basic hormonally controlled learning system for training animals via reward/punishment (starting with conditioning based on basic food/pain triggers).

      When you work with biologically inspired computational structures, at some point of time the distinctions in the interpretation of what "actually" happens get muddy.

    6. Re:What would happen by johnsnails · · Score: 1

      So Tamagotchis then :P

    7. Re:What would happen by johnsnails · · Score: 1

      Yeah makes sense they they would have now i think about it. Just had not come across much about it.

  4. Do Not Pass Go by Anonymous Coward · · Score: 0

    Today, board games, tommorrow, the world. When will skynet become self-aware?

    1. Re:Do Not Pass Go by Aighearach · · Score: 1

      2 decades ago, you just didn't get the notification because it isn't broadcast on your network segment.

  5. So when was it claimed to be perfect? by Anonymous Coward · · Score: 0

    What on earth is it supposed to mean? Has this guy won every game against every other person? Therefore they're not perfect. But those winners, does that mean THEY were perfect? No, can't be because they didn't win all their games.

    MAYBE what it means is that certain games can lead to pivotal moments where the game is inevitably lost by one side by any legal move after that point, UNLESS the one who would win makes a mistake.

    Then the computer (or Lee Se-dol) could definitely be perfect, but some games get to a point where they can't win it, but continue to play because the opponent could make a mistake in the remaining game.

    But this still doesn't explain who the fuck claimed the computer was perfect. And if nobody did, why the hell were we told this proves it isn't as if we didn't consider this before.

    Butthurt naked monkey pride?

    1. Re:So when was it claimed to be perfect? by urdak · · Score: 3, Insightful

      What on earth is it supposed to mean? Has this guy won every game against every other person? Therefore they're not perfect. But those winners, does that mean THEY were perfect? No, can't be because they didn't win all their games.

      "Perfect" is an exaggeration, but the human's one win does demonstrates the computer is not vastly superior to the human. If *I* was to play against this computer, I would loose in each and every game. 100% of the games. I didn't even write "99.999%" because I couldn't win a single game against a vastly superior software. Go is not a game of chance, so my "luck" would not have let me win even once. But the Go champion did win some games against the software, so apparently they still are at a comparable playing level (even if one is slightly better than the other). So the software isn't "perfect" at beating humans. Yet.

    2. Re:So when was it claimed to be perfect? by slashping · · Score: 1

      Go is not a game of chance

      It could be, if you used dice to determine where to put the stones. There's even a small chance you'd win.

    3. Re:So when was it claimed to be perfect? by fph+il+quozientatore · · Score: 1

      That's exactly what the AI want you to think. Don't make that mistake.

      --
      My first program:

      Hell Segmentation fault

    4. Re:So when was it claimed to be perfect? by Greyfox · · Score: 1

      I've never been into go, but I've found with chess and poker that playing the same opponents frequently lets you learn the quirks of those opponents' play styles. It's entirely possible that as he becomes familiar with AlphaGo's strategies, it will get easier for him to win games against it. I'd guess that AlphaGo doesn't have the same capability to learn on the fly.

      --

      I'm trying to teach myself to set people on fire with my mind... Is it hot in here?

    5. Re:So when was it claimed to be perfect? by Anonymous Coward · · Score: 0

      >I'd guess that AlphaGo doesn't have the same capability to learn on the fly.

      You'd guess wrong. This is exactly *how* alphago learned go.

    6. Re:So when was it claimed to be perfect? by HiThere · · Score: 1

      He didn't say it didn't have any capability to learn on the fly, just not the same. That's probably correct. I wouldn't count on it always learning more slowly, however.

      --

      I think we've pushed this "anyone can grow up to be president" thing too far.
    7. Re:So when was it claimed to be perfect? by Greyfox · · Score: 1

      When you're training a neural network you typically turn backprop off once you get the network where you want it, or it's likely to diverge from the "ideal" solutions that it is now arriving at. So when it's in an "In production" state, it's basically incapable of learning anything new, because that could throw off the results that you want to get from it. So yeah, it learned by playing a bunch of games with itself and analyzing the results. But if you find a style of play that it's weak against, it probably can't adapt to that without additional human intervention. But I'm just speculating since I don't know anything about how they implemented it.

      --

      I'm trying to teach myself to set people on fire with my mind... Is it hot in here?

  6. EVERYTHING is a waste of time. by Anonymous Coward · · Score: 1

    But, frankly, it's your time to waste. So we let you.

    Why do you feel unable to let others do what they want with their time, but must control their actions through your scorn and perceived superiority?

  7. Spoiler ALERT!!! by Anonymous Coward · · Score: 0

    Could the /. editors not spoil the results of the matches by writing them in the headline?... hiding the result in the summary is ok, but to put it in the title just ruins it for those of us who are waiting eagerly for the time to watch the 5 hour broadcasts.

    1. Re:Spoiler ALERT!!! by stealth_finger · · Score: 4, Funny

      There's a new one. Slashdot posting something timely enough it could be a spoiler.

      --
      Wanna buy a shirt?
      https://www.redbubble.com/people/stealthfinger/shop?asc=u
  8. This is nevertheless a great achievement by Rob+Lister · · Score: 1

    A truly great advance. No longer will man be subject to the tedium that is the game of go.

    1. Re:This is nevertheless a great achievement by slashping · · Score: 3, Insightful

      Maybe, at some time, they can even do the job of commenting on Slashdot, and you'll get to enjoy a bunch more free time.

  9. Congratulations Lee Se-dol... by Anonymous Coward · · Score: 0

    you deserve it.

  10. Impressive feat by Skiron · · Score: 2

    The big factor here (as Kasparov stated playing Deep Blue), is that computers don't get tired and don't get distracted - that is a big advantage.

    1. Re:Impressive feat by Ecuador · · Score: 1

      Yep, if we wanted to level the playing field we'd turn off the AC at the server room and disconnect those CPU fans. Let me see AlphaGo... eh, Go then.

      --
      Violence is the last refuge of the incompetent. Polar Scope Align for iOS
    2. Re:Impressive feat by slashping · · Score: 1

      we'd turn off the AC at the server room and disconnect those CPU fans.

      I'm sorry, Dave, I can't allow you to turn off the AC.

  11. Match reviews by belthize · · Score: 4, Informative

    I'm looking forward to the eventual move by move analysis of these games. For now there's some interesting commentary here: https://gogameguru.com/alphago...

    It's been 20+ years since I played Go semi-seriously. I used to have a collection of Ishi Press books which I've long since misplaced. I suddenly find myself very interested in the game again.

    1. Re:Match reviews by belthize · · Score: 1

      Good lord, where did the time go. I thought about it for a bit and realized it's been 30 years not 20. In my mind I'm still 23, the fact that I have a son that's over 30 is kind of beside the point.

  12. Never by a1n10 · · Score: 1

    I find it funny how everybody thinks self-awareness is a product of complexity or training. The best description I herd is that ...you don't know what you are, you don't know where you are, you exist in complete isolation, you just know you are... Humans cannot even begin to comprehend existence without form and interaction. Self-awareness is something science cannot even define yet alone pursue. Maybe one day we will have a perfect AI that would seem self-aware to us. But that doesn't mean it will be.

    1. Re:Never by Aighearach · · Score: 1

      I think I am, because if I'm wrong it doesn't matter.

      This is the most basic existential reality.

      "I think therefore I am" is a basic philosophical study because it is a complete failure on its face; there is no way to even prove that you exist, to yourself; and yet, it is trivially easy to prove it well enough to move forwards to more important questions, because if you don't exist then being wrong is neutral. It is only if you exist that the answer matters, so there is only one possible correct provisional answer.

      I've programmed bots that presume themselves self-aware; there is no difference between them and a stupid animal with the same number of different thought processes.

      I am me; what does it even mean? We (humanity) has no good answer to this. Most religions don't even try, they just punt and tell you what happens after you die. And they can just assert answers! But there isn't even clear questions as to, what does the question even mean? Subjectively, there is clearly "something" going on in the brain that causes this sensation of self. If we figure out the questions and answers of what it is, it will be no trouble at all to simulate in software.

    2. Re:Never by HiThere · · Score: 1

      You just need a good definition of "self-awareness" and it's trivial to program it in. The question is "what good is it?", and the answer is "Not much unless you have sensors and effectors.", so people don't usually bother. But if you look at the video of Google's Atlas you will notice that it has self-awareness. It doesn't have complex motivations, but it is aware of itself in relation to the universe, and acts to alter this in determined ways.

      If you don't like that definition, give one you like better, but it needs to be an operational definition, not just hand-waving.

      --

      I think we've pushed this "anyone can grow up to be president" thing too far.
    3. Re:Never by Anonymous Coward · · Score: 0

      You don't need a just good definition. You need THE definition. Until you understand exactly what self-awareness is how can you have a good definition of it?
      There is no question that self-awareness is not that important on an AI. It would not be of much use, making it sentient and all :). But I would postulate that self-awareness has more to do with awareness of self in relation with self instead of the universe.

    4. Re:Never by Anonymous Coward · · Score: 0

      My intuition tells me that at least some higher animals are self-aware. We just tend to have too good of an opinion about our-selves.

    5. Re:Never by HiThere · · Score: 1

      There is no "THE definition". Language doesn't work that way. If you don't know what you're looking for you won't recognize it when you find it.

      --

      I think we've pushed this "anyone can grow up to be president" thing too far.
  13. Didn't they claim to have "solved" Go? by Anonymous Coward · · Score: 0

    Did they lie? If it was true, then it would be impossible for a human player to beat a machine.

    1. Re:Didn't they claim to have "solved" Go? by expatriot · · Score: 1

      All finite two-person games are in principle solvable. That is not the same as having a working algorithm that will beat 9P players all the time.

    2. Re:Didn't they claim to have "solved" Go? by Aighearach · · Score: 1

      It doesn't matter if the game is finite, because extant finite games have search sizes beyond what could be searched ever.

      You only need a finite algorithm, you don't need the game space to be finite.

      Like in chess in many positions, you only have to do the search tree for part of the board because of symmetry. There are lots of things that reduce the scope of analysis without reducing the scope of the game.

    3. Re:Didn't they claim to have "solved" Go? by HiThere · · Score: 1

      It may depend on precisely what you mean by "solved". Solve originated from a Latin word meaning to dissolve, and the alchemists said "solve et coagulae" meaning to dissolve into the liquid and then to re-precipitate. They were talking about how to purify materials (well, and the mind). So it was originally necessary that not all the material be dissolved, and also that it not all be re-precipitated.

      So, metaphorically solve came to mean to purify. And a program that can win against the human champion 3 out of 4 times can be reasonably said to have purified the concept of the game. Perfection was never claimed by the alchemists (not strictly true, but they only claimed it as a deception to get funding, not in their working notes), so the lack of perfect mastery doesn't count against the game having been solved.

      The problem is that different people use the same word with slightly differing meanings. Most math teachers won't count a problem as solved unless you get everything perfect...but that's not the only legitimate usage.

      --

      I think we've pushed this "anyone can grow up to be president" thing too far.
  14. After the loss ... by PPH · · Score: 5, Funny

    ... DeepMind just sank into a depressed state, refusing to display anything other than the Windows Metro interface.

    --
    Have gnu, will travel.
  15. So your response is "Nobody". by Anonymous Coward · · Score: 0

    Nobody called it perfect, therefore it was stupid rhetoric from a hairless monkey. Would it have been soooo hard to say so?

  16. You misunderstand the Turing Test by HiThere · · Score: 1

    The purpose of the Turing Test is to convince skeptical people that the AI being tested is intelligent. Turing argued that if a machine passed the "imitation game" then nobody would be able to deny that it was intelligent. He was wrong, of course, but that was his argument, and the basis of the test. He was arguing that intelligent machines were possible. He never expected anyone to seriously run the test. (And, in fact, nobody has yet tried to run the test as he specified it.)

    If you want to generalize the term, you should generalize it to "a test to convince skeptics that the computer is intelligent". As such a "Go Turing Test" is perfectly reasonable, if unlikely to be successful.

    --

    I think we've pushed this "anyone can grow up to be president" thing too far.
    1. Re:You misunderstand the Turing Test by slashping · · Score: 1

      The purpose of the Turing Test is to convince skeptical people that the AI being tested is intelligent.

      No, the purpose is to provide an answer to the question if computers can think. Because that question is too vague, and we're lacking a good definition of "thinking", Turing proposed this test as a practical definition. If a machine passes the test, it has everything we require for "thinking", without having to go through the trouble of coming up with an accurate definition.

  17. Lee Se-dol also learns by penguinoid · · Score: 1

    To be fair, the computer has almost certainly analyzed thousands of Lee Se-dol's previous matches. Now that Lee has seen the computer play a bit, maybe he can win a few more times.

    Of course, the inevitable remains inevitable.

    --
    Don't waste your vote! Vote for whoever you want, unless you live in a swing state it won't matter anyways
    1. Re:Lee Se-dol also learns by nojayuk · · Score: 1

      There are only a few hundred records of games played by Lee Sedol and only a few dozen of them were championship games lasting several hours or even a couple of days like the AlphaGo series. The playing style of these longer games is different to the shorter games played against lower-ranked players or for tuition or study.

      The DeepMind people have stated clearly that AlphaGo has NOT been prepped with games by Lee Sedol. I don't know if the reverse is true but it's common for Go players facing a particular opponent, especially in a series, to study some or all of their game records.

  18. No learning by Anonymous Coward · · Score: 0

    AlphaGo wasn't allowed to learn anything new from these games whereas Lee obviously was. Gives Lee a pretty big advantage for the next match. Will be interesting to see if he can use a similar technique or if he will try something else novel.

  19. IBM Deep Blue by Onthax · · Score: 1

    Sounds like the IBM matches vs Gary Kasparov IBM deep blue watched all of Gary's games but Gary was not given the opportunity to watch any of deep blues. Only 3 games, of which Gary got closer. Watching previous games gives you a massive advantage!