Sekėjai

Ieškoti šiame dienoraštyje

2012 m. gruodžio 27 d., ketvirtadienis

What computers can do and how?

The answer is: still not much, but significantly more than earlier.

_If you follow these things, you know that the field of artificial intelligence has for decades been a frustrating underachiever, as it proved incredibly hard for computers to do things every human being finds easy, like understanding ordinary speech or recognizing different objects in a picture.

Lately, however, the barriers seem to have fallen — not because we’ve learned to replicate human understanding, but because computers can now yield seemingly intelligent results by searching for patterns in huge databases.

True, speech recognition is still imperfect; according to the software, one irate caller informed me that I was “fall issue yet.” But it’s vastly better than it was just a few years ago, and has already become a seriously useful tool. Object recognition is a bit further behind: it’s still a source of excitement that a computer network fed images from YouTube spontaneously learned to identify cats. But it’s not a large step from there to a host of economically important applications[1]._
We are going to see soon what those implications are.
1. PAUL KRUGMAN. Is Growth Over?

Komentarų nėra: