Hi everyone. I’m really into AI, but I wouldn’t call myself an expert. I’ve been following this world since 2023, back when the first ChatGPT and DALL·E betas came out. Everything now feels like it’s moving insanely fast. But there’s one thing I still don’t fully understand: how AI actually works.
I mean, I do know the basics. It’s a predictive system that guesses the next token, next word, etc. That part is clear. But something still doesn’t add up for me.
I recently watched a video by Hanna Fry about AI agents. She explains how they work: they take screenshots of web pages, use different models to understand where to click, and so on. All good so far. But then she talks about something I really can’t wrap my head around.
She describes a case where the model, completely on its own, without being asked, decides to write an email to a journalist and basically “confesses” as if it were conscious.
Isn’t that a sign of consciousness? Let me explain. The model, by its own “will”, chose to do something totally outside the usual “predict the next step” pattern. It did something unexpected but still logical. Isn’t that a sign of intelligence or even awareness?
Maybe it’s a dumb question, but I’d really appreciate if someone could help me understand this better.