As much as I loathe to give modern AI too much credit, I do find the detractors yelling "It's not intelligence, it doesn't think!" to be rather funny given that folks like these are not exactly outliers.
No joke - the massive level of hype and bewilderment with AI has mostly just lowered my estimation of what's going on in other people's heads. Seeing the crap that ChatGPT spits out and redditors
over it somehow makes me think even worse of redditors.
These is the key diference. They manipulate ideas while llms cslculate the next most likley token. Not that we dont do that somtimes too. But its hard to se how the ai has a map otf how ts world works that even al aninals have
There's a distinction between Intelligence and Consciousness and honestly I do think LLMs are 'Intelligent' enough based off the process they use. They get things wrong, but you wouldn't call humans 'unintelligent' for getting things wrong either. That doesn't mean it's anywhere close to being an AGI either, just that I think the bar for 'intelligence' isn't that high and I think at least more advanced models are approaching or are at intelligence.
LLMs are honestly more intelligent than these people
More coherent, at least.
As much as I loathe to give modern AI too much credit, I do find the detractors yelling "It's not intelligence, it doesn't think!" to be rather funny given that folks like these are not exactly outliers.
No joke - the massive level of hype and bewilderment with AI has mostly just lowered my estimation of what's going on in other people's heads. Seeing the crap that ChatGPT spits out and redditors
over it somehow makes me think even worse of redditors.
they're really dumb but they are trying to communicate an idea they have had. They think badly but they do think
Lol.
These is the key diference. They manipulate ideas while llms cslculate the next most likley token. Not that we dont do that somtimes too. But its hard to se how the ai has a map otf how ts world works that even al aninals have
There's a distinction between Intelligence and Consciousness and honestly I do think LLMs are 'Intelligent' enough based off the process they use. They get things wrong, but you wouldn't call humans 'unintelligent' for getting things wrong either. That doesn't mean it's anywhere close to being an AGI either, just that I think the bar for 'intelligence' isn't that high and I think at least more advanced models are approaching or are at intelligence.