Hacker News Reader

GPT-3 is a bit more than a Markhov chain. There is a lot more to it than just being big. Same for GPT-2.

As soon as you put the intelligence in AI this changes. It is easy to say that current AI systems are “dumb” (although note that the precise meaning of dumb has changed significantly over the last few decades). You can say that about any AI with sub-human intelligence. But if you reach human levels, you can likely reach super-human as well, so you need to start worrying much earlier.

GPT-3 is better than everything what was achieved so far in multiple metrics, though yes, it's of course not an "intelligence".

However, this is ignoring possible future developments - things may get better and more advanced and that's what AI-safety people are concerned about.