2021年经济学人 自动驾驶汽车(3)(在线收听

You could, for instance, ask a car fitted with a reasoning engine why it had hit the brakes, and it would be able to tell you that it thought a bicycle hidden by a van was about to enter the intersection ahead.

例如,如果问一辆装有推理引擎的汽车为什么踩刹车,它会告诉你它认为有一辆藏在货车后面的自行车即将进入前面的十字路口。

A machine-learning program cannot do that.

而机器学习程序无法做到这一点。

Besides helping improve program design, such information will, Dr Bhatt reckons, help regulators and insurance companies.

巴特博士认为,这些信息除了有助于改进项目设计之外,对监管机构和保险公司也会起到帮助作用。

It may thus speed up public acceptance of autonomous vehicles.

因此,它可能会加快公众对自动驾驶汽车的接受程度。

Dr Bhatt's work is part of a long-standing debate in the field of artificial intelligence.

巴特博士的工作在人工智能领域长期备受争论。

Early AI researchers, working in the 1950s, chalked up some successes using this sort of preprogrammed reasoning.

20世纪50年代的早期人工智能研究人员使用这种预编程推理取得了一些成功。

But, beginning in the 1990s, machine learning improved dramatically, thanks to better programming techniques combined with more powerful computers and the availability of more data.

但是,从20世纪90年代开始,由于更好的编程技术、更强大的计算机以及更多数据的可用性,机器学习得到了显著改善。

Today almost all AI is based on it.

今天,几乎所有的人工智能都基于机器学习。

Dr Bhatt is not, though, alone in his scepticism.

不过,巴特博士并不是唯一持怀疑态度的人。

Gary Marcus, who studies psychology and neural science at New York University and is also the boss of an AI and robotics company called Robust.AI, agrees.

加里·马库斯在纽约大学研究心理学和神经科学,他也是一家名为Robust.AI的人工智能和机器人公司的老板。

To support his point of view, Dr Marcus cites a much-publicised result, albeit from eight years ago.

为了支持他的观点,马库斯博士引用了一个8年前被广泛报道的结果。

This was when engineers at DeepMind (then an independent company, now part of Google) wrote a program that could learn, without being given any hints about the rules, how to play Breakout, a video game which involves hitting a moving virtual ball with a virtual paddle.

当时,DeepMind(当时是一家独立公司,现在隶属于谷歌)的工程师编写了一个程序,可以在不需要任何规则提示的情况下学习如何玩Breakout,这是一款电子游戏,需要用虚拟球拍击打移动中的虚拟球。

DeepMind's program was a great player.

DeepMind的程序是一个杰出的玩家。

But when another group of researchers tinkered with Breakout's code—shifting the location of the paddles by just a few pixels—its abilities plummeted.

但是,当另一组研究人员对Breakout的代码进行修改,将电板的位置稍稍改变了几个像素后,它的能力便一落千丈。

It was not able to generalise what it had learned from a specific situation even to a situation that was only slightly different.

它无法将自己从某个特定情境中学到的东西应用到即便只是略微不同的情形中。

For Dr Marcus, this example highlights the fragility of machine-learning.

对马库斯博士来说,这个例子凸显了机器学习的脆弱性。

But others think it is symbolic reasoning which is brittle, and that machine learning still has a lot of mileage left in it.

但也有人认为符号推理才是脆弱的,而机器学习还有很大的进步空间。

Among them is Jeff Hawke, vice-president of technology at Wayve, a self-driving-car firm in London.

伦敦的自动驾驶汽车公司Wayve的技术副总裁杰夫·霍克就是其中之一。

Wayve's approach is to train the software elements running a car's various components simultaneously, rather than separately.

Wayve的方法是对驱动汽车各个部件的软件同时开展训练,而不是分别训练。

In demonstrations, Wayve's cars make good decisions while navigating narrow, heavily trafficked London streets—a task that challenges many humans.

在演示中,Wayve的汽车在狭窄、拥挤的伦敦街道上行驶时做出了正确的决定——这对于人类驾驶员来说往往也颇具挑战。

  原文地址:http://www.tingroom.com/lesson/2021jjxr/534989.html