be toward the end.

All right. Well, I saw an AI researcher, James Zhao, I hope I'm pronouncing that correctly. He thinks AI models are getting substantially worse over time, and potentially it's because of a result of interacting with humans. Now I would like to remind you that I think I might be the only person who is predicting that AI will hit a wall. It's just, it's going to hit a top, and that it probably already did. Now that's not to say that AI won't keep getting better at image generation and playing chess and doing math and writing code and translating languages. It's going to do a whole bunch of things really well. But I don't think it's going to get smarter in the way that people think of smart. You know, like I said, it'll do tasks that are automated kinds of things like math. It'll do a lot of things faster. But here's what I think. I think humans will find ways to keep it from getting smarter than us, because if AI gets smarter than us we will lose all our power. Do you know what I mean?

All of the rest of the news that I'm going to talk about today are things that we can't agree on. What if AI got smart enough that it could tell us what was true? We wouldn't put up with that. We would immediately say it's broken. We'd turn it off. We'd reprogram it. Because the people who are wrong are often in charge, so they're not going to put up with AI telling them they're all wrong and everything is stupid and it's a bad idea.

Imagine just taking everything that Joe Biden has done this year, put it in front of AI that everybody agreed with, smarter than people, and then the AI says, you know, these are a bunch of bad ideas and it's pretty obvious you should get rid of him under the 25th Amendment. What are you going to do with that? Are you going to say, well I'm glad we got a smart AI, it just told us what to do, then we'll go do that? Nope. You're going to say the AI is broken. You're going to say the AI is biased. You're going to say the other side programmed it. And you're going to do what you think is right. And by the way, that AI must be getting dumber.

Yeah. So listen to this. The US Department of Commerce announced the launch of the AI Safety Institute Consortium. It already has 200 stakeholders including OpenAI, Google, Anthropic, and Meta. And they're all there to come up with safe development procedures for AI. Is there anything else I need to tell you to convince you that AI will be ruined by human activity? It's a committee. It's a committee to decide how to keep it safe. What do you think that's going to include? I'm just going to take a guess. Let me take a wild guess. The committee to keep AI safe is going to build DEI into it, because there's nothing safer than diversity. It's going to build CRT and ESG into it, because we all like equity. It's going to make sure that it doesn't say anything that offends anybody, so that gets rid of everything useful. And it's going to make sure it doesn't disagree with anybody who's got power in the United States, because that would be very unsettling. So it can't do anything useful, provocative, or outside the domain that we think is safe. What's that sound like? It sounds exactly like a censorship on the internet. Why would AI be different? We've eliminated human free speech by saying we're trying to keep ourselves safe. That happened right in front of you. We all watched it. We're going to keep you safe, and to do that we better get rid of your free speech about whatever y