If you’re talking about LLMs, then you’re judging the tool by the wrong metric. They’re not designed to solve problems or pass captchas - they’re designed to generate coherent, natural-sounding text. That’s the task they’re trained for, and that’s where their narrow intelligence lies.
The fact that people expect factual accuracy or problem-solving ability is a mismatch between expectations and design - not a failure of the system itself. You're blaming the hammer for not turning screws.
Consciousness - or “self-awareness” - has never been a requirement for something to qualify as artificial intelligence. It’s an important topic about AI, sure, but it’s a separate discussion entirely. You don’t need self-awareness to solve problems, learn patterns, or outperform humans at specific tasks - and that’s what intelligence, in this context, actually means.
You’re describing intelligence more like a soul than a system - something that must question, create, and will things into existence. But that’s a human ideal, not a scientific definition. In practice, intelligence is the ability to solve problems, generalize across contexts, and adapt to novel inputs. LLMs and chess engines both do that - they just do it without a sense of self.
A calculator doesn’t qualify because it runs "fixed code" with no learning or generalization. There's no flexibility to it. It can't adapt.
People should really be more specific when they're speaking of AI. I'm assuming you mean generative AI in the form of video clips.
Sure, why not? We've pretty much accepted awful CGI effects as the norm already. I hardly see AI generated effects as any worse than that. It's entirely possible that those effects could even be better.
Both. It originated in a city with a lab doing gain-of-function research on coronaviruses. I’m not claiming that’s definitely where it came from - but it’s quite the coincidence, to say the least.
And there’s nothing wrong with anti-China propaganda as long as it’s aimed at the authoritarian government, not the people living under it. That regime deserves every bit of it.
“But what about the US this and the UK that?!” Yeah - they deserve it too.
No. Reddit was better in almost every possible way with just few exeptions. I'm simply too principled to go back and I see Lemmy as the only viable alternative.
I’m pretty happy having four distinct seasons. I don’t like winter and snow at all, but I think suffering through six months of cold and darkness is exactly why the warmth and sunshine feel so damn good when summer finally comes. Also, with climate change, the climate where I live has - so far - only been getting better. I’m not saying it’s good overall, but it’s not all bad either.
I mostly downvote bad faith, hostile or just generally angry/mean comments independent of whether I agree with them or not. Basically I'm using it against the people that are polluting the air here.
If you’re talking about LLMs, then you’re judging the tool by the wrong metric. They’re not designed to solve problems or pass captchas - they’re designed to generate coherent, natural-sounding text. That’s the task they’re trained for, and that’s where their narrow intelligence lies.
The fact that people expect factual accuracy or problem-solving ability is a mismatch between expectations and design - not a failure of the system itself. You're blaming the hammer for not turning screws.