Post

Demonstrating reasoning and language capabilities in computers alters our understanding of self, even if we don't particularly like the alteration

This is the crux of the debate, right? Whether it's doing mimicry or reasoning. I struggle to see how you could view it as purely mimicry. It has to tie together multiple novel thoughts in a logical construction to create its responses.

This is what concerns me about AI boosters. They see it as a simplistic dichotomy, mimicry or reasoning? It is neither. My favorite example is a prompt like “what is your favorite color?”. The mimic result would be the most popular color. The reasoning result would be

“I don’t know what colors mean”. Instead it churns the tokens and comes up with an answer based on what it calculates the reply you want is. I’m other words it is neither reasoning or mimicry. It is how the engine is coded

I'm going to guess it says something playful like "I don't really have a preference, but" or "I'm just an agent that doesn't have favorites" and then something like "but if I had to choose, _ is a lovely color." Let's see...

0 reposts0 likes