投稿者: 
科学と工学
以前の投稿で, ニューラルネットワークにおける言語習得は人間のそれとは異なると述べるチョムスキーの主張を見た. 今回はさらに一歩踏み込み, この点に関して明確にEngineering (工学)と言い切っている主張をより深く見たい.
I don’t see any point to them at all. But what is the point of a system that can match 45 tera-things in 45 terabytes of data and make you think you are listening to something real. That is a game. But as long as AI is following that direction, it may be useful for linguistics, like maybe it can study huge number amounts of data and find properties that were not noticed, something like that. It could be a good tool. But again, it is an engineering result. And in another area it also does interesting things like by studying enormous number of ways in which the proteins can fold. AI methods have partially answered some very interesting questions of molecular biology about protein folding, or the things like that. (Chomsky, 2022, p. 364)
錯覚させるシステムはゲームであり, それはEngineering (工学)の結果である. この観点が現在持たれているAI研究の中でも明確な立ち位置を表明している一文であると言えるであろう.
実に生成文法らしい立場である.
参考文献
- Chomsky, N. (2022). Genuine explanation and the strong minimalist thesis. Cognitive Semantics, 8(3), 347-365.