Agree. I'll also say it's a lot further away than your comment implies. Once ChatGPT4 can add metonymy by itself in a random discussion, talking about consciousness in silicon can be interesting.
It was only a few years ago that AlphaGo beat the best human at GO. Supposedly a game so deep that only some unknown quality of being a human could ever grasp it.
And AlphaStar also wins against humans in SC2, where there is hidden information and 'deception'.
I'd say that is more worry than GPT.
GPT is a language model.
AlphaGo is winning in a competitive game.
and
Life is just a game.
Just start adding more variables, and more goals.
Start combining GPT/AlphaGo/Visions Systems/etc.... It might not be that far.
