The puzzle solver returns lol, been a while.
I tested three LLMs the other day with a few sentences with all lowercase words and capital letters strewn about and asked if it could spell out what the capital letters said. Gemini, ChadGPT and Copilot, they all failed spectacularly, I even tried helping but it was like talking to a shitcoiner.
Meanwhile Elon suing OpenAI saying they have AGI capabilities.
Just give us the AI to make droids to help (and protect) humans in the physical world.
Anytime I hear anyone say anything about AGI
Like, there's no freakin way we're anywhere close to AGI, right? I'm very open to being corrected here, but like, at the moment I'm not seeing it.
@OneOneSeven, thanks for the welcome back
reply
Good point, I had to look up what it's supposed to mean from the guy who coined it:
The definition I use now is that [AGI] is a system that can do all the cognitive kinds of tasks that people can do, and possibly more"
Pretty vague and general, but definitely far from where the current LLMs are now, per my example. But who knows what goes on behind closed doors at ClosedAI.
reply