Stories
Slash Boxes
Comments

SoylentNews is people

posted by martyb on Thursday August 09 2018, @10:03PM   Printer-friendly

Submitted via IRC for SoyCow1984

And it wasn't even close.

A month and a half ago, OpenAI showed off the latest iteration of its Dota 2 bots, which had matured to the point of playing and winning a full five-on-five game against human opponents. Those artificial intelligence agents learned everything by themselves, exploring and experimenting on the complex Dota playing field at a learning rate of 180 years per day. [...] the so-called OpenAI Five truly earned their credibility by defeating a team of four pro players and one Dota 2 commentator in a best-of-three series of games.

There were a few conditions to make the game manageable for the AI, such as a narrower pool of 18 Dota heroes to choose from (instead of the full 100+) and item delivery couriers that are invincible. But those simplifications did little to detract from just how impressive an achievement today's win was.

[...] play-by-play commentator Austin "Capitalist" Walsh sums up the despondency felt by Team Human after the bout neatly:

Never felt more useless in my life but we're having fun at least so I think we're winning in spirit.

Sure aren't winning in-game

— Cap (@DotACapitalist) August 5, 2018

Source: https://www.theverge.com/2018/8/6/17655086/dota2-openai-bots-professional-gaming-ai

Dota 2 is a sequel to Defense of the Ancients (DotA).

Previously: OpenAI Bot Bursts Into the Ring, Humiliates Top Dota 2 Pro Gamer in 'Scary' One-on-One Bout
OpenAI to Face Off Against Top Dota 2 Players in 5v5 Match-ups


Original Submission

Related Stories

OpenAI Bot Bursts Into the Ring, Humiliates Top Dota 2 Pro Gamer in 'Scary' One-on-One Bout 34 comments

In the past hour or so, an AI bot crushed a noted professional video games player at Dota 2 in a series of one-on-one showdowns.

The computer player was built, trained and optimized by OpenAI, Elon Musk's AI boffinry squad based in San Francisco, California. In a shock move on Friday evening, the software agent squared up to top Dota 2 pro gamer Dendi, a Ukrainian 27-year-old, at the Dota 2 world championships dubbed The International.

The OpenAI agent beat Dendi in less than 10 minutes in the first round, and trounced him again in a second round, securing victory in a best-of-three match. "This guy is scary," a shocked Dendi told the huge crowd watching the battle at the event. Musk was jubilant.

OpenAI first ever to defeat world's best players in competitive eSports. Vastly more complex than traditional board games like chess & Go.

— Elon Musk (@elonmusk)

According to OpenAI, its machine-learning bot was also able to pwn two other top human players this week: SumaiL and Arteezy. Although it's an impressive breakthrough, it's important to note this popular strategy game is usually played as a five-versus-five team game – a rather difficult environment for bots to handle.

[...] It's unclear exactly how OpenAI's bot was trained as the research outfit has not yet published any technical details. But a short blog post today describes a technique called "self-play" in which the agent started from scratch with no knowledge and was trained using supervised learning over a two-week period, repeatedly playing against itself. Its performance gets better over time as it continues to play the strategy game. It learns to predict its opponent's movements and pick which strategies are best in unfamiliar scenarios.

OpenAI said the next step is to create a team of Dota 2 bots that can compete or collaborate with human players in five-on-five matches. ®

Youtube Video

Also covered here (with more vids, including the bout in question):
Ars Technica: Elon Musk's Dota 2 AI beats the professionals at their own game
Technology Review: AI Crushed a Human at Dota 2 (But That Was the Easy Bit)
TechCrunch: OpenAI bot remains undefeated against world's greatest Dota 2 players


Original Submission

OpenAI to Face Off Against Top Dota 2 Players in 5v5 Match-ups 3 comments

AI bots trained for 180 years a day to beat humans at Dota 2

Beating humans at board games is passé in the AI world. Now, top academics and tech companies want to challenge us at video games instead. Today, OpenAI, a research lab founded by Elon Musk and Sam Altman, announced its latest milestone: a team of AI agents that can beat the top 1 percent of amateurs at popular battle arena game Dota 2.

You may remember that OpenAI first strode into the world of Dota 2 last August, unveiling a system that could beat the top players at 1v1 matches. However, this game type greatly reduces the challenge of Dota 2. OpenAI has now upgraded its bots to play humans in 5v5 match-ups, which require more coordination and long-term planning. And while OpenAI has yet to challenge the game's very best players, it will do so later this year at The International, a Dota 2 tournament that's the biggest annual event on the e-sports calendar.

[...] OpenAI says that at any one time its Dota 2 bots have to choose between 1,000 different actions while processing 20,000 data points that represent what's happening in the game.

[...] For this new batch of Dota bots, the amount of self-play is staggering. Every day, the bots played 180 years of game time at an accelerated rate. They trained at this pace over a period of months.

Dota 2 is a sequel to Defense of the Ancients (DotA).

Previously: OpenAI Bot Bursts Into the Ring, Humiliates Top Dota 2 Pro Gamer in 'Scary' One-on-One Bout


Original Submission

OpenAI Develops Text-Generating Algorithm, Considers It Too Dangerous to Release 49 comments

New AI fake text generator may be too dangerous to release, say creators

The creators of a revolutionary AI system that can write news stories and works of fiction – dubbed "deepfakes for text" – have taken the unusual step of not releasing their research publicly, for fear of potential misuse.

OpenAI, an nonprofit research company backed by Elon Musk, says its new AI model, called GPT2 is so good and the risk of malicious use so high that it is breaking from its normal practice of releasing the full research to the public in order to allow more time to discuss the ramifications of the technological breakthrough.

At its core, GPT2 is a text generator. The AI system is fed text, anything from a few words to a whole page, and asked to write the next few sentences based on its predictions of what should come next. The system is pushing the boundaries of what was thought possible, both in terms of the quality of the output, and the wide variety of potential uses.

When used to simply generate new text, GPT2 is capable of writing plausible passages that match what it is given in both style and subject. It rarely shows any of the quirks that mark out previous AI systems, such as forgetting what it is writing about midway through a paragraph, or mangling the syntax of long sentences.

More like ClosedAI or OpenAIEEEEEE.

Related: OpenAI 'Universe' Platform Provides Means to Link Image Recognition Vehicular AI Agents Into GTA 5
The OpenAI Dota 2 Bots Defeated a Team of Former Pros


Original Submission

This discussion has been archived. No new comments can be posted.
Display Options Threshold/Breakthrough Mark All as Read Mark All as Unread
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
(1)
  • (Score: 0) by Anonymous Coward on Thursday August 09 2018, @11:13PM (1 child)

    by Anonymous Coward on Thursday August 09 2018, @11:13PM (#719676)

    This article has sparingly little meaning as-is. What are we paying the editors for?

    • (Score: -1, Spam) by Anonymous Coward on Thursday August 09 2018, @11:21PM

      by Anonymous Coward on Thursday August 09 2018, @11:21PM (#719681)

      Quickly, quickly! There was no time to waste. They had to get everything ready before next month, or it would be too late. Yes, these three men firmly believed this to be the case; they firmly believed that the apocalypse was imminent.

      The apocalypse. To them, it was only 35 days away, and so their haste was understandable. This was why they were rapidly stockpiling supplies in their underground bunker in preparation for the end of the world. Would they be able to meet their goals? They can and they did, with just a week to spare. But upon doing this, something went terribly wrong.

      The supplies were simply too delectable. At first, it was just one. Then, it was another, then another, and then yet more. The men continued playing with the toys and breaking them, all the while convincing themselves that just one more can't hurt anything. It did not take long before they had expended all of their supplies. And the apocalypse, they believed, was only a day away.

      They couldn't believe it. How had this happened? How!? They had captured over 200 cute, young children to violate after the apocalypse occurred, but now they were all utilized and beginning to rot. The men wondered, 'Why were we so greedy!?' The men feared, 'How will we survive the end of the world now!?' The men cried, 'Why are children so fragile!?' They wept, cried, and screamed, but the corpses of the expended children simply continued to rot.

      Some would call the men foolish. Others would understand what had transpired, since ignoring such delicacies would be very difficult. But regardless of which view one holds, one thing was certain: The men were tragic victims. After all, their intent was simply to survive what they believed to be the end of the world, and yet all those children broke far too easily. Without a doubt, the men's fundamental rights had been violated. It was a tragedy of the highest order.

      But then these men - these kind-hearted men - were given another chance. You see, the apocalypse they so firmly believed would happen never actually occurred. At first they were surprised and confused, but then their surprise and confusion became relief. Their calculations had been off all along. The apocalypse, they now believed, would actually occur in two months. The men smiled to themselves, their hope restored and their hearts strong.

      It was time to begin gathering supplies...

  • (Score: 2) by archfeld on Thursday August 09 2018, @11:13PM (2 children)

    by archfeld (4650) <treboreel@live.com> on Thursday August 09 2018, @11:13PM (#719677) Journal

    So a group of 'self' programmed aim bots win at a controlled slice of a game ? I wonder how they would do at a real scenario open ended game like a WoW boss battle or such. I'd assume they would eventually learn by trial and error each step but would they figure out the best way and the better pay-outs ? I still debate the title AI and see them as bots, but the technology is getting better. Some day they will make the entire point of the game moot and kill the already sickened golden goose. What will the game companies do then ?

    --
    For the NSA : Explosives, guns, assassination, conspiracy, primers, detonators, initiators, main charge, nuclear charge
    • (Score: 3, Interesting) by Gaaark on Friday August 10 2018, @12:05AM (1 child)

      by Gaaark (41) on Friday August 10 2018, @12:05AM (#719698) Journal

      I imagine it would it be kind of like asking a chess champion to play a game where he can only move his pawns?
      The AI might play a better game just because it's learned to play that way, while the champ would have to relearn how to play restricted chess.

      --
      --- Please remind me if I haven't been civil to you: I'm channeling MDC. ---Gaaark 2.0 ---
      • (Score: 2) by darkfeline on Friday August 10 2018, @06:19PM

        by darkfeline (1030) on Friday August 10 2018, @06:19PM (#720013) Homepage

        That's what they always say. It was only a few years between an AI (pre-machine learning advances) beating a Go player with a significant handicap to beating a low dan Go player evenly, then a few month to beating the best human player. Especially with machine learning, AI has been shown to improve faster and faster.

        Today the AI beats former pros at a limited version of the game, next year they'll beat everyone.

        --
        Join the SDF Public Access UNIX System today!
  • (Score: 2, Funny) by Anonymous Coward on Friday August 10 2018, @12:59AM (2 children)

    by Anonymous Coward on Friday August 10 2018, @12:59AM (#719713)

    Finally, I can buy video games and enjoy them even if I don't have time to play them. I'll simply buy my console's AI DLC and let it play for me.

    [10 minutes later]

    Wow, I'm so much better than I used to be. I really like this game.

  • (Score: 2) by snufu on Friday August 10 2018, @02:08AM (3 children)

    by snufu (5855) on Friday August 10 2018, @02:08AM (#719742)

    For every example of a first time occurrence of AI being able to do something better than humans, we are reminded of numerous examples of trivial human tasks where the field of AI is no where near solving.

    What are some games that will not fall to AI for some time?

  • (Score: 2) by rigrig on Friday August 10 2018, @03:33AM (1 child)

    by rigrig (5129) Subscriber Badge <soylentnews@tubul.net> on Friday August 10 2018, @03:33AM (#719766) Homepage

    So General AI isn't quite here yet, but at least we're throwing our narrow AI computing resources at things that benefit mankind, like curing(or at least diagnosing) cancer.
    Sorry, I meant stuff that benefits shareholders(like predicting stock markets), as proper shareholders only believe in short-time ROI (and never get cancer)
    Oh, wait...
    Uhm....
    Increasing productivity by proving yet another game isn't worth your time/effort, because computers can play it better anyway?

    --
    No one remembers the singer.
  • (Score: 2) by Gaaark on Friday August 10 2018, @09:10PM

    by Gaaark (41) on Friday August 10 2018, @09:10PM (#720070) Journal

    Yup, but for now, it might have been a handicap for the real players.

    Makes you wonder if there will be any South Koreans earning the six figure incomes in the future when they can be easily beat by a program.
    I'd be looking for a new job if i were them.

    --
    --- Please remind me if I haven't been civil to you: I'm channeling MDC. ---Gaaark 2.0 ---
(1)