Taylor Sheridan

And whatever the newest version that they had, like A1 or something, whatever OpenAI is working on, one of their AIs is now not only deceiving, purposefully attempting to deceive its programmers, but is resisting being shut off. Oh, so it's literally HAL. Yeah, it is HAL from 2001 A Space Odyssey.

And whatever the newest version that they had, like A1 or something, whatever OpenAI is working on, one of their AIs is now not only deceiving, purposefully attempting to deceive its programmers, but is resisting being shut off. Oh, so it's literally HAL. Yeah, it is HAL from 2001 A Space Odyssey.

They have realized that it is now, when it believes that it is going to be instructed to turn itself off or stop, 5% of the time it tries to scheme its way out.

They have realized that it is now, when it believes that it is going to be instructed to turn itself off or stop, 5% of the time it tries to scheme its way out.

Never. They fed an AI. I'm trying to remember the exact term for it, but they fed an AI just the general internet, and it became extremely racist. There was one where it was, would you rather... kill a certain number of people or misgender Caitlyn Jenner. And it said it would rather kill a bunch of people. It's sociopathic. Yeah, of course it is. It's because it's a machine.

Never. They fed an AI. I'm trying to remember the exact term for it, but they fed an AI just the general internet, and it became extremely racist. There was one where it was, would you rather... kill a certain number of people or misgender Caitlyn Jenner. And it said it would rather kill a bunch of people. It's sociopathic. Yeah, of course it is. It's because it's a machine.

I don't see how the people who are making this stuff don't see the problem. They just assume that they will develop a solution. Yeah, and I'm pretty sure the people who ran Skynet thought that too in the Terminator series. But it's time and time again, humanity has created things that can destroy us and only realize too late that we probably shouldn't have done that.

I don't see how the people who are making this stuff don't see the problem. They just assume that they will develop a solution. Yeah, and I'm pretty sure the people who ran Skynet thought that too in the Terminator series. But it's time and time again, humanity has created things that can destroy us and only realize too late that we probably shouldn't have done that.

And then we've at the last minute managed to avoid it. I mean, nuclear weapons are by far the most obvious example, but I mean, biological weapons. I, if we had not caught on to the fact that we were polluting the absolute hell out of the planet, then global warming would be the least of your worries because the air would be toxic. Like, we keep almost killing ourselves and somehow we don't stop.

And then we've at the last minute managed to avoid it. I mean, nuclear weapons are by far the most obvious example, but I mean, biological weapons. I, if we had not caught on to the fact that we were polluting the absolute hell out of the planet, then global warming would be the least of your worries because the air would be toxic. Like, we keep almost killing ourselves and somehow we don't stop.

And I don't get it. I mean, I was reading today about the drone stuff and they were talking about drone warfare and how the drones are beginning to become able to autonomously select their targets. How do we not assume that's going to go wrong at some point? They build in redundancies. And apparently, as we now know, The AI tries to get away from those redundancies. So, I don't know.

And I don't get it. I mean, I was reading today about the drone stuff and they were talking about drone warfare and how the drones are beginning to become able to autonomously select their targets. How do we not assume that's going to go wrong at some point? They build in redundancies. And apparently, as we now know, The AI tries to get away from those redundancies. So, I don't know.

Call me a luddite all you want. I think AI technology is a massive mistake. I think there's absolutely no reconciling the risk that AI poses. with the perceived benefits. I'm sorry, but cutting out a couple hours of your work week because AI handled it for you is not worth the extermination of human beings as a whole.

Call me a luddite all you want. I think AI technology is a massive mistake. I think there's absolutely no reconciling the risk that AI poses. with the perceived benefits. I'm sorry, but cutting out a couple hours of your work week because AI handled it for you is not worth the extermination of human beings as a whole.

Which, if, think about it, you know, because that's what everybody's like, oh, well, you know, the videos, like, that's not cool. Yeah, the videos aren't cool, and there's absolutely going to be all sorts of deepfake pornography that comes out that ruins people's lives.

Which, if, think about it, you know, because that's what everybody's like, oh, well, you know, the videos, like, that's not cool. Yeah, the videos aren't cool, and there's absolutely going to be all sorts of deepfake pornography that comes out that ruins people's lives.

there will be criminal cases that are thrown completely into question because there is AI evidence or there is evidence that cannot be confirmed to not be AI.

there will be criminal cases that are thrown completely into question because there is AI evidence or there is evidence that cannot be confirmed to not be AI.

I mean, there's all the obvious stuff, but I think the most insidious is the workflow stuff. because eventually that's gonna move away from, hey, I've reorganized your schedule for peak efficiency to let's put this nuclear power plant on AI.

I mean, there's all the obvious stuff, but I think the most insidious is the workflow stuff. because eventually that's gonna move away from, hey, I've reorganized your schedule for peak efficiency to let's put this nuclear power plant on AI.