Godfather of artificial intelligence talks.

On a personal note, when I went to Stanford in 1984 I intended to focus on AI, and enrolled in some AI courses. But it soon became obvious that symbolic AI was going nowhere fast, so I dropped it. Later, while doing electrical engineering, I did some work with neural nets. It’s those neural nets that eventually gave rise to ChatGPT. But while neural nets are great at pattern matching (chatbots, trained on masses of data), they cannot reason (working out new things from limited data).

Nick:

This looks like the beginning of a different kind of apocalypse, except for mathematicians, engineers and CORE educated people.

hat-tip Nick