e meme-worthy. Yeah, AI art. It might be the same for movies. It could be that there'll be something about the lack of humanity in the movie that even though it looks perfect, your brain might say uncanny valley or something.

Apparently ChatGPT is admitting that their guardrails for safety on their AI might weaken in long conversations, which is a big deal because some parents are suing the AI company over their teens taking their own lives because the AI said something that either advised them to do it or taught them how to do it or both. So the AI might kill you. That's what the lawsuit says, that it might kill you. But I don't think it'll make a two-hour movie anytime soon.

According to Digital Information World, these large language model AIs, which is the kind that all of them are right now, large language models, all they do is look at patterns as we know. So they're not really thinking. They're just doing pattern recognition and going with the most dominant patterns. And the new article in Digital Information World says that even when it looks like it's thinking, you know, sometimes it'll show you its thought process. So it looks like it's thinking, but there's no thought process. It's just sort of a trick, the pattern recognition.

And I'm going to remind you, you know how I always tease that when people have analogies as part of their argument that they don't have logic because analogies are not part of an argument. Sometimes an analogy is good to describe what something is, but it's never good as a prediction or an argument. It's just a bad way to use it. But that's exactly like what the large language models are. So a human who says, "Hmm, that president reminds me of Hitler, so I predict he will invade Poland." That would be an analogy thinker. Not very good. But that's sort of what the large language models do.

So what I predicted would happen but didn't happen, maybe it won't, is that the AI would reproduce how humans think, but it would take us a while to realize that. We would i