Newman steered a course back to the safer waters of mathematics, pointing to the act of imagination that had been required to connect the 'real numbers' of length with the integers of counting, which involved 'seeing analogies between things that had not been put together before.
纽曼则保守地讨论了一些数学问题,他指出,长度是实数,而计数则是整数,将这两者联系起来,需要进行一种联想:"这种联想就是在两个事物之间找出一种关系。
Can we even guess at the way a machine could make such an invention from a programme composed by a man who had not the concept in his own mind?
那么一个机器程序,能否在编程者不事先设定的情况下,自动地做出这样的联想?
Alan could guess, in fact; it was just the kind of thing he was thinking about:
实际上这正是图灵所想的事情:
I think you could make a machine spot an analogy, in fact it's quite a good instance of how a machine could be made to do some of those things that one usually regards as essentially a human monopoly.
我认为机器是可以进行联想的,而且这本身是一个很好的例子,体现机器如何做到一些我们认为是人类专利的事情。
Suppose that someone was trying to explain the double negative to me, for instance, that if a thing isn't not-green it must be green, and couldn't quite get it across.
假设说,有人给我讲解"双重否定",比如说"这个东西不是非绿色的",那么它就是绿色的,这个东西不容易直接讲清楚。
He might say,'Well, it's like crossing the road. You cross it, and then you cross it again, and you're back where you started.' This remark might just clinch it.
但他可以说:"好吧,这就像你过马路,你过一次,然后再过一次,就会回到原来的地方。"这样的说法,就能解释这个问题。
This is one of the things one would like to work with machines, and I think it would be likely to happen with them.
现在人们希望,机器也能做到这样的类比,而我认为,机器是可以做到的。