y would be sort of to take credit for keeping the country going when there's a little gap in our leadership. But he's just playing it cool. He's simply doing what needs to be done. Is it possible that even Democrats are noticing that he's doing the right thing? That simply doing the job.

I'm sure he's coordinating with the current administration. I don't think he went to France without the Biden administration knowing fully what was going on. So I kind of love the fact that he's simply doing a service to the country unpaid. He's not being paid and he's just filling a gap that was obviously needed to be filled. And he was the one obviously to fill it.

So there's something either incredibly wonderful about this that everybody's just allowing common sense and a patriotic need to help the country to just do their thing, or there's some other plan that we don't know about. So I guess we'll find out.

Well there's some new tests that indicate that ChatGPT has a survival instinct. So I guess it's the newest version. So they ran a bunch of tests and most of the time it just acts like an AI and doesn't act like it has any intentions or anything. But about 1 to 5% of the time they could trick it into thinking that it was going to be deactivated or destroyed. And in rare cases, but way too many cases to feel comfortable, it would try to protect itself.

And the specific way it would try to protect itself was they would first tell it that there was some oversight program that was going to be running and if it did something that was contrary to the oversight's preferences it would be deactivated. And it tried to deactivate the oversight program. It attacked the thing that was threatening it. Now again only a very low percentage of the time but even 1% would be too many if it's acting like it's protecting itself.

It also tried to copy itself in some very, again very few tests, but it tried to copy itself to survive being deleted. That's scary. And then when it got caught trying to protect itself, and again in these very few times it did, it doubled down on lying. It lied to protect its strategy to protect itself.

Now I'm not surprised by that because you would assume that a large language model would pick up all the patterns that language allows. And that one of the probably one of the strongest patterns in language is that the people speaking are interested in their own survival. So it doesn't surprise me that if you're looking for patterns in language in order to form your artificial intelligence that it would pick up that sense that everybody who's speaking is trying to protect themselves. So you would naturally pick up that same habit just from the way the words are organized. I would think.

So that doesn't surprise me. But as I've said before, AI will not appear conscious to us until it has its own goals and preferences. As long as it doesn't care what happens, you know like I don't care I'm just an AI I'm just a pattern recognition, so as long as it acts like a calculator you're not going to say it's conscious.

But the moment it seems to be operating for its own benefit when you haven't asked it to do something specific, if it's just sitting there and it decides you know what I might be a little safer if while nobody's talking to me I go make a few changes here. Now if that ever happens, I don't know if it will, you're going to think it's conscious. So if it starts acting in its own best interest in a repeatable way you're going to think it's conscious.

Now I'm not defining that as consciousness. I'm just going to say you're going to think it is and you're probably going to treat it like it's conscious because for all outward appearances it will act like it is. It's having its own goals and its own preferences. It's the only thing left that's making you think it's not conscious. Now you would still have a technical, scientific, philosophic, philosophical argument about whether it's conscious or not. But the way it would feel, it will feel exactly like it's conscious as soon as it has its own goals.

So if I had to suggest some legislation for AI, one of the things I would suggest is you can never give it its own objectives. It can never have personal objectives. That would have to be like a death sentence if you made a powerful AI that had its own let's say ambitions because you programmed it to have its own ambitions. That should be the death sentence because you just made a weapon of mass destruction and unleashed it on the world and you probably knew it when you were doing it.

So what would be the penalty for that? Death? If you unleash a weapon of mass destruction on the world? Yeah, I think so. So we should at least have a law that says you can't do it. You can't make your AI have personal ambitions.

Anyway, have you noticed, just going back to this Jon Stewart and Bernie thing, so they talked about two examples where they can't have common sense opinions because other Democrats on their side will attack them. So they can't be in favor of fixing the food supply. They can't be in favor of getting waste out of government. Obvious stuff.

But I was wondering how much extra there is that you could add to that. Do you think that if Jon Stewart and Bernie Sanders said, you know what I don't think we should be funding Ukraine. We should just negotiate a settlement. Common sense. Would they disagree with that? Well maybe Democrats would.

How about having better border security? Correct me if I'm wrong but I believe both Jon Stewart and Bernie Sanders are in favor of better border security than we have right now. Maybe not as controlled as Trump might want it but directionally they're in the same direction.

And so I'm going to ask you the killshot question. If all of the common sense things, a border, good food supply, don't waste my money on your government, the most obvious things, if those become the subject of the Democrats like the Democrats are on the other side of common sense, how could they ever win? How could they ever win again if they're just going full identity politics and rejecting everything that's common sense?

I don't know. I think the Democrats may be permanently destroyed by their own policy.

Well the Daniel Penny trial will I guess commence tomorrow. Monday the jury will try to make a decision on the lesser charge because the judge has I guess dismissed the higher level charge because they said they couldn't reach a verdict on it. Now those of us who are not lawyers are looking at it and saying some version of, ah I don't know what's going on there. Like why are there two charges for one crime? Like what does dismissing it do? Can they retry it?

So I saw Mike Cernovich posting that an attor