I don't know how many people have the absolute position that it "can't be used as a tool to improve anything".

There's a lot of "it's risky and expensive and anti-labor, so I don't think we should be pouring all this effort into it"

They're expressing a sentiment, not making a factual statement.

Ofcourse AI can improve things - it already has. Even LLMs have their use cases. However while improving some things it can meanwhile also make other things much worse.

I think you’re in denial.

I think a lot of those people have one specific idea of what AI is, and don't realize how they are making machines more energy efficient, and analyzing patient data to cure diseases. It's a tool that is being used for both good and evil.

I think most people don't know what AI is. AI has been improving things for decades. Most people assume AI means LLMs now. I think they suck in many ways, but to deny that they have zero value is dumb.

When this development phase of ai slows down you’ll end up with as many or more model types than there are people on this planet, so just like with people, some models will be great resources and valuable tools, some will be excellent time wasters, some will be terrible, some will be malicious, etc

Shhhh, don't disturb the hivemind.