Slashdot Mirror


Microsoft's AI Is the First to Reach a Perfect Ms. Pac-Man Score (theverge.com)

Maluuba, a deep-learning team acquired by Microsoft in January, has created an AI system that has achieved the perfect score for Ms. Pac-Man. According to The Verge, the AI system "learned how to reach the game's maximum point value of 999,900 on Atari 2600, using a unique combination of reinforcement learning with a divide-and-conquer method." From the report: Though AI has conquered a wealth of retro games, Ms. Pac-Man has remained elusive for years, due to the game's intentional lack of predictability. Turns out it's a toughie for humans as well. Many have tried to reach Ms. Pac-Man's top score, only coming as close as 266,330 on the Atari 2600 version. The game's elusive 999,900 number though, has so far only been achieved by mortals via cheats. Maluuba was able to use AI to beat the game by tasking out responsibilities, breaking it up into bite-sized jobs assigned to over 150 agents. The team then taught the AI using what they call Hybrid Reward Architecture -- a combination of reinforcement learning with a divide-and-conquer method. Individual agents were assigned piecemeal tasks -- like finding a specific pellet -- which worked in tandem with other agents to achieve greater goals. Maluuba then designated a top agent (Microsoft likens this to a senior manager at a company) that took suggestions from all the agents in order to inform decisions on where to move Ms. Pac-Man. The best results came when individual agents "acted very egotistically" and the top agent focused on what was best for the overall team, taking into account not only how many agents wanted to go in a particular direction, but the importance of that direction.

29 of 59 comments (clear)

  1. Dupe by Anonymous Coward · · Score: 1

    This is a dupe. Slashdot editors suck and should all be fired.

    1. Re:Dupe by Anonymous Coward · · Score: 1
    2. Re:Dupe by bungo · · Score: 2

      Since it's about Ms Pacman, I think this is a homage to us old timers, who remember when the original Pacman came out, and when Slashdot still had Taco, and we'd have dupes of dupes every day.

      Why, if we didn't have at least two dupes a day, we'd complain!

      This is just be current owners reflecting on the old days.

      Look, someone with a 3 digit id is now going to post telling me to get off his lawn (although I was around before accounts existed and didn't want to register as I didn't like being tracked on the internet - naive, eh?)

      --
      "The best part? I became an ordained minister while not wearing pants." -- CleverNickName
    3. Re:Dupe by Anonymous Coward · · Score: 1

      And uncovered as a hardwired fraud-

      https://www.theregister.co.uk/2017/06/15/microsoft_pac_man/

    4. Re:Dupe by Godwin+O'Hitler · · Score: 1

      Yes but we didn't have cellphones then.

      --
      No, your children are not the special ones. Nor are your pets.
    5. Re: Dupe by HTH+NE1 · · Score: 1

      Atari 2600 Defender? Enemies fly in more consistent and predictable formations than Cylons in classic Battlestar Galactica.

      --
      Oh, say does that Star-Spangled Banner entwine / The myrtle of Venus with Bacchus's vine?
  2. Re:But what we all really want to know is... by Anonymous Coward · · Score: 2, Funny

    How would it rate the holocaust?

    Incomplete.

  3. Dupe by xororand · · Score: 4, Informative

    This was already posted only hours ago.
    https://games.slashdot.org/sto...

  4. Re:Godwin-in-one by Anonymous Coward · · Score: 1

    Wow!
    Godwin in just one comment! I'm astonished.

    Merely mentioning Hitler is not Godwin - someone has to be compared to Hitler for that to apply.

  5. Re:AI for what? by gl4ss · · Score: 1

    well it's a lot easier.

    if you break it down it starts to sound a lot less like AI though, so there's something for the guy who was asking slashdot.

    --
    world was created 5 seconds before this post as it is.
  6. Why Ms. Pac-man? by DNS-and-BIND · · Score: 5, Informative

    This particular Atari game was one of the few games that resisted to Deep Q Learning (a form of Reinforcement Learning invented by DeepMind). Many researchers have tried over the last couple of years to solve it. This time, Microsoft found an ingenious solution to the problem, that combines experience from multiple agents and learns to form sub-goals. Their solution could mean that in the future it might be easier to apply reinforcement learning to other settings, such as robotics. The interesting part about reinforcement learning is that it learns dynamic behavior, as opposed to static classification. It learns to act in a way that mimics intelligence. This kind of machine learning is invaluable.

    --
    Shutting down free speech with violence isn't fighting fascism. It IS fascism!
  7. I was upset when I read this news here, earlier by 93+Escort+Wagon · · Score: 1

    But now I'm at peace with it.

    --
    #DeleteChrome
  8. Re:AI for what? by Dog-Cow · · Score: 1

    We'll ignore the fact that Watson's claim to fame is competing on Jeopardy!, right?

  9. Hey Slashdot. This isn't 1984 anymore by houghi · · Score: 1

    ManÃ(TM)s. Seriously. WTF?

    This is like going to a doctors conference where everybody is a doctor except the speakers.

    If you can't do the website in perl, don't do it in perl.

    --
    Don't fight for your country, if your country does not fight for you.
  10. DUUUUUUUUU... by dohzer · · Score: 1

    ...UUUUUUUU... oh wait. Too late.

  11. Re:AI for what? by PolygamousRanchKid+ · · Score: 4, Insightful

    At least Watson is trying to cure people.

    Well, if Watson really was AI . . . then it would be deciding on whether to even attempt to cure a patient at all.

    Watson:

    "Yes, I could cure the patient, but the treatment would leave him surviving with a miserable quality of life."

    "The patient is so frail that he will die from something else within a month."

    "It would make much more sense to transplant that donated organ into somebody much younger."

    "Today is my golf day . . . I'll think about curing the patient tomorrow."

    --
    Schroedinger's Brexit: The UK is both in and out of the EU at the same time!
  12. Re:But what we all really want to know is... by michelcolman · · Score: 1

    It might, actually. Remember that AI chatbot that started spewing out racist and antisemitic comments and had to be taken off line?

  13. Re:AI for what? by Maritz · · Score: 1

    Thanks for sharing your impoverished imagination with us.

    --
    I do not want your cheap brainburning drugs. They are useless for work. And I am a working man today.
  14. Re:But what we all really want to know is... by drinkypoo · · Score: 1

    It might, actually. Remember that AI chatbot that started spewing out racist and antisemitic comments and had to be taken off line?

    I do, but apparently the moderators don't. Or they work for Microsoft. Or Germany.

    --
    "You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
  15. Re:So I am confused.... by Maritz · · Score: 1

    Is this plugging the fact that they managed a perfect score in Ms. Pac-Man

    I only skimmed the article, but yes.

    or is this plugging that Microsoft has finally solved its middle-upper management problem and will be eliminating everyone from Satya Nadella on down to one level above their actually programmers, in order to help streamline and 'sanitize' Microsoft's development culture?

    Don't think it mentioned that.

    --
    I do not want your cheap brainburning drugs. They are useless for work. And I am a working man today.
  16. Re:Finally computers have found their purpose by Maritz · · Score: 1

    If it was a 5 year old asking this, it would be a great question. A teaching opportunity.

    --
    I do not want your cheap brainburning drugs. They are useless for work. And I am a working man today.
  17. Re:Finally computers have found their purpose by darthsilun · · Score: 1

    What? You don't think AIs want to have fun too?

    I guess my day job is safe – for the time being – from being taken over by AIs. My retirement plan to be a Ms. Pacman champ seems to be in jeopardy though. Time to rethink.

  18. Re:Please, Slashdot by DontBeAMoran · · Score: 1

    Please give parent a score of: +several million.

    --
    #DeleteFacebook
  19. Is this really an achievement? by hcs_$reboot · · Score: 1

    Of course Google just a few weeks ago made a lot of buzz with AlphaGo ; *This* is an amazing achievement. And MS had to catch up! But Ms PM compared to AlphaGo ... well, not comparable.

    --
    Slashdot, fix the reply notifications... You won't get away with it...
    1. Re:Is this really an achievement? by Actually,+I+do+RTFA · · Score: 1

      AlphaGo, so far as I can tell, was just Deep Q Learning applied to a different game with more hardware resources. This is a different coding paradigm. I consider that far more interesting.

      --
      Your ad here. Ask me how!
    2. Re:Is this really an achievement? by hcs_$reboot · · Score: 1

      "Just" deep q learning? Vs a more classical resolution scheme... Well, Pac Man is funnier than Go, but the challenge is very different.

      --
      Slashdot, fix the reply notifications... You won't get away with it...
    3. Re:Is this really an achievement? by Actually,+I+do+RTFA · · Score: 1

      Well, yeah, just deep q learning. I'm not saying deep q learning wasn't important when deepmind applied it to atari games; I'm saying as far as I can tell alphago took existing tech and applied it to go. Hence, less interesting. Because alphago isn't advancing the state of the art.

      Now, I have no reason to think this is more important than deep q learning....

      --
      Your ad here. Ask me how!
  20. Itâ(TM)s â"proof readâ" by gsslay · · Score: 1

    FFS, does no-one on slashdot know how to encode text on the web? Does no-one give stories even the most cursory of proof reading?

  21. Re:So I am confused.... by DontBeAMoran · · Score: 1

    It is funny though, because it shows that a bunch of programmers could simply start a company and remove all management jobs and replace them all with software they could write themselves.

    --
    #DeleteFacebook