Whatever you do, don’t put it on Jupiter! It’ll get more stupider!

Maybe we should put datacenters on Mars, to warm it up.

Civilian homes are not tactical targets.

It doesn’t have access to the training data. It’s not outputting training data, it’s making up something that feels like the training data.

Really? Must be a new feature, it didn’t when I tried it. I know they can execute code, I guess the engineers added a search tool. Regardless, that tool isn’t part of their fundamental design. It’s something they have to go and access, and most of the time they won’t. If you were to experiment by asking it to write a scientific paper, you’d find the references are garbage with broken links and nonexistent papers. Hallucinations. It’s just making something plausible sounding up, the same as a lazy human might.

if they actually could “think” wouldnt they factcheck themselves first before saying something

No. They don’t have access to the original training data, or to the internet. They’re stuck remembering it the same way a human remembers something: with neurons. They cannot search the dataset for you. The best they can do is remember and tell you.

No, LLMs are not just an interface for accessing training data. If that were true, then their references would actually work. The fact that LLMs can hallucinate and make stuff up proves that they are not just accessing the training data. The ANN is generating new (often incorrect) information.

Their only condition was that they wanted to keep the Emperor. It was ridiculous of the Allies to demand a wholly conditional surrender. All those people got blown up just to win the argument about that one point. They could have ran a conventional air bombing campaign against tactical targets, but they decided to drop nukes on a “tactical” target in the middle of a huge city! And then they did it again! That’s not tactical, that’s strategic. If you’re going to use nukes, at least use them on a military base far away from cities.

Forcing people to give their picture to Peter Thiel is not regulation. Age verification laws are giving companies more licence to abuse users, not less.

Models aren’t programs.

So you’re saying that because the AI has been exposed to training data in the past, it’s incapable of making choices. Interesting argument. Pretty easy to reducto ad absurdum, though.

No human has ever deployed tactical nukes against a nuclear capable enemy.

The Japanese government was already willing to surrender.