有没有人认为目前的人工智能方法会遇到瓶颈？
数十亿美元的投入，以及对几年内实现通用人工智能（AGI）的巨大宣传。是否有人认为目前基于深度学习/神经网络的人工智能方法最终会走入死胡同，无法兑现其承诺？如果是的话，为什么？
我意识到这个问题的定义有些模糊。毫无疑问，目前的方法将继续改进并取得成果，因此可能不容易界定“死胡同”。
在这个背景下，我想看看是否有人认为当前的方向是错误的，并且不会带我们到达最终目标。
查看原文
Billions of dollars spent, incredible hype that we will have AGI in several years. Does anyone think the current deep learning / neural net based AI approach will eventually hit a dead end and not be able to deliver its promises? If yes, why?<p>I realize this question is somewhat loosely defined. No doubt the current approach will continue to improve and yield results so it might not be easy to define "dead end".<p>In the spirit of things, I want to see whether some people think the current direction is wrong and won't get us to the final destination.