Blog

  • Google’s AI Experiments help you understand neural networks by playing with them

    Google’s work in machine learning and artificial intelligence is often interesting, but it can be a bit academic. People like to get their hands on these things — as much as you can, anyway, with something intangible. To that end, Google is collecting a bunch of little demonstrations of this emerging category of tech in…

  • Data, A Human Composer And Artificial Intelligence Create A Symphony In The Louvre

    Facebook recently announced a new app that uses artificial intelligence (AI) to make live videos look like art. In September 2016, AI was used to judge a beauty contest called Beauty.Ai 2.0 with controversial results, primarily that the AI selected mostly contestants with white skin. That same month on September 20, 2016, AI entered the Louvre in Paris and was the center of…

  • Adobe intros AI framework Sensei, teases Project Felix for Creative Cloud

    Adobe on Wednesday introduced two new platforms that bring machine learning, artificial intelligence and 3D image design to its cloud portfolio. First up, the creative software giant rolled out Sensei, which it describes as “a framework and set of intelligent services built into the Adobe Cloud Platform.” Adobe says it’s been using machine learning and…

  • AI judge predicts outcome of human rights cases with remarkable accuracy

    An artificial intelligence algorithm has predicted the outcome of human rights trials with 79 percent accuracy, according to a study published today in PeerJ Computer Science. Developed by researchers from the University College London (UCL), the University of Sheffield, and the University of Pennsylvania, the system is the first of its kind trained solely on…

  • Here’s how close AI is to beating humans in different games

    Two decades after Deep Blue conquered chess and a year after  AlphaGo took down go, there are still games that artificial  intelligence can’t beat.   AI researchers like games because they offer complex,  concrete, and exciting challenges, which can unlock broader  applications. Take  it from AlphaGo creator Demis Hassabis: Games  “are useful as a testbed, a platform for trying…

  • Public Input and Next Steps on the Future of Artificial Intelligence

    As part of the White House Future of Artificial Intelligence Initiative, OSTP is releasing the public comments submitted on artificial intelligence, sharing insights from five events across the country over the past months, and announcing a new White House event on artificial intelligence at the Frontiers Conference in October, 2016. In June, the White House…

  • Google Brain’s neural-net AI dreams up its own encryption strategy

    It’s fun to write about developments in artificial intelligence like they’re harbingers of an impending AIpocalypse. Jokes about our new robot overlords notwithstanding, computers are getting scary smart these days, and it’s not always flattering to compare humans with AI. The machines can outperform humans in a lot of important ways: we routinely trust robot surgeons,…

  • 机器智能的崛起 计算机国际象棋

    就在19年前,当 IBM的超级计算机深蓝色击败Garry Kasparov 时,就实现了AI世界的一个里程碑。 在此之前,他是不败的世界国际象棋冠军,这可能是有史以来最伟大的人类球员。 这是AI简短历史上的重要事件。 自1970年代以来,计算机国际象棋程序就一直在下棋,并提高了他们的比赛水平将击败绝大多数人口。 我自己记得在1980年代初购买国际象棋计划,该计划提供了从初学者到高级的6个级别的比赛。 即使那样,我还是很难击败高于3级的机器。到卡斯帕罗夫(Kasparov)发挥深蓝色时,国际象棋游戏软件的质量正在迅速提高。 但是,包括卡斯帕罗夫本人在内的大多数专家都认为击败大师的步伐是非常不可能的。 比赛于1997年5月在纽约举行,涉及六场比赛中的最佳成绩。 卡斯帕罗夫(Kasparov)赢得了第一场比赛,但在第二场比赛中意外击败。 卡斯帕罗夫(Kasparov)显然对这场失败感到震��,第二天的新闻发布会上,他指责深蓝色作弊。 他通过声称表现出不可预测的行为来理解这一点,他认为这是由于IBM编程团队在比赛中篡改所致。 规则规定,程序员可以更改游戏之间的程序,但在游戏期间不会更改程序。 IBM团队抓住了卡斯帕罗夫(Kasparov)措手不及,因为他相信计算机国际象棋程序虽然快速且计算机上无瑕,但由于其可预测的敷衍了事的行为而不会宣称大师的头皮。 在Kasparov在第一场比赛中击败了Deep Blue之后,IBM团队在软件中产生了更多随机的不可预测性。 它起作用了,深蓝色继续赢得比赛。 直到这场失败,卡斯帕罗夫一直有理由对机器智能的限制进行一些理由。 对于深蓝色,本质上使用了AI技术,当时涉及“蛮力”搜索以在国际象棋中获胜。 蛮力搜索是AI 早期的常用范式,它试图通过迅速通过数百万的动作组合来迅速搜索到具有计算机力量的对手 – 在深蓝色的情况下,分析了超过2亿个可能的动作,每秒 。 使用修剪方法通常会减少搜索空间(即可能的移动)。 这很重要,因为在国际象棋比赛中,球员通常仅限于每举动三分钟的时间。 但是,任何人都无法在一生中分析2亿可能的举动,更不用说一秒钟了。 但这对当时的卡斯帕罗夫来说并不重要,因为他认为人类的智慧和多年经验使他具有直觉的见解,而他不需要分析。 确实,当他曾经被问到他每秒分析多少动作时,他宣称:“不到一个”。 这意味着当时的战线是在愚蠢机器的卓越计算能力和人类大师的创意,有见地的天才之间广泛绘制的。 但是19年了,AI世界发生了很大变化。 如今,正如卡斯帕罗夫(Kasparov)本人所承认的那样:“一款运行免费的国际象棋计划的体面笔记本电脑将粉碎深蓝色和任何人类的祖母。象棋机器的跳跃是可预测和弱的,到了可怕的强者,只花了十二年 ”。 卡斯帕罗夫(Kasparov)似乎已成为一个convert依,现在识别了计算机国际象棋对人类国际象棋群众的好处的见解和发现。 他为什么现在这么说? 因为计算机硬件继续保持不懈的速度,但AI程序也不再像AI初期那样依靠蛮力搜索算法。 如今,语言翻译程序或无人驾驶汽车和高级国际象棋程序的AI使用技术,例如遗传算法和神经网络 – 更类似于人类智能的工作方式。 这些技术提供的是以前的技术所没有的,这既是匹配模仿人类思维的模式的能力,也可以学习的能力。 优秀的人类国际象棋参与者,例如其他主题领域的专家,使用根据经验建立的模式识别技能,而AI技术现在变得擅长于模式匹配 – 直到最近,许多人认为这不太可能。 学习技术可以改善国际象棋软件并将其提高到新的水平。 据说人类进化的关键里程碑之一是时间,估计是100万年前,当时我们的灵长类动物祖先是通过观察他人在工作中学到的。 达到了这一点,花了数十亿年的生物进化。 但是,现在许多人认为,在未来几十年中,AI计划将获得与人类相同的学习能力水平。 这确实令人惊讶,并提出一个问题,AI将我们带到哪里? 我将在下一篇文章中进一步讨论。

  • Japanese AI Writes a Novel, Nearly Wins Literary Award

    I had thought my job was safe from automation–a computer couldn’t possibly replicate the complex creativity of human language in writing or piece together a coherent story. I may have been wrong. Authors beware, because an AI-written novel just made it past the first round of screening for a national literary prize in Japan. The…

  • Obama: My Successor Will Govern a Country Being Transformed by AI

    President Obama thinks that artificial intelligence will be one of the thorny issues awaiting his successor in the White House. The implications of recent advances in AI were a major theme at a technology conference organized by the White House and held in Pittsburgh today. Speaking on stage, Obama said self-driving Ubers, which are being…