0:00 AI would never do anything bad, right? It would never lie to you, right, Brad? Former OpenAI CEO Sam Altman was ousted by the rest of the company's non-profit board on Friday. A key character in the spectacle has been OpenAI chief scientist and board member Ilya Sutskiver. Sutskiver has established himself as an esoteric spiritual leader at the company. The Atlantic's new details are bizarre, even by the standards of tech industry wackadoos. Feel the AGI! Feel the AGI! Employees reportedly chanted, per The Atlantic, a refrain that was led by Sutskiver himself. The chief scientist even commissioned a wooden effigy to represent an unaligned AI that works against the intrusions.
0:55 A new version of OpenAI's There is a real sense in which we are creating something very different from anything we've created so far. Yeah, I mean... we definitely will be able to create completely autonomous beings with their own goals. And it will be very important, especially as these beings become much smarter than humans, it's going to be important to have these beings, the goals of these beings be aligned with our goals. I feel that technology is a force of nature. I feel like there is a lot of similarity between technology and biological evolution. It is very easy to understand how biological evolution works. You have mutations, you have natural selections, you keep the good ones, the ones that survive. And just through this process you're going to have huge complexity in your organisms. We cannot understand how the human body works because we understand evolution, but we understand the process more or less. And I think machine learning is in a similar state right now, especially deep learning. We have very simple, a very simple rule that takes the information from the data and puts it into the model. And we just keep repeating this process. And as a result of this process, the complexity from the data gets transformed, transferred into the complexity of the model. So the resulting model is really complex and we don't really know exactly how it works. You need to investigate. But the algorithm that did it is very simple. The goals of these beings being aligned with ours, it's almost as if we're, I mean, we're actually living in a place that we are essentially playing God. He doesn't believe this because he believes in evolution from a biblical perspective. He's saying, we have to make sure that these things don't sin. We have to make sure that they're on the same playing field, that they're abiding by the laws of love. And if they're not, that's gonna be very scary for us, because it's dangerous. So..
3:43 Let me, let me take that. Can I take that? and just move it one little yard further down the field. Okay. So you're saying that they must be—it's as if how God looks at us and says they have to be in line and not sin against me. Well, he's talking about AI, and he specifically is saying that he wants it in line with human interests. What does the Bible say about the heart of man, Brad? Yeah, the heart of man is wicked, and who can know it? So I think it's interesting that he's saying we need to—you remember? iRobot, Will Smith, they had a thing in there where they had to obey the three laws. Yeah, I know. The three laws, your perfect circle of protection. A robot cannot harm a human being. The first law of robotics. Yeah, I know. I've seen your commercials. But doesn't the second law state that a robot has to obey any order given by a human being?
4:34 What if it was given in order to kill? Impossible. It would conflict with the first law. Right, but the third law states that a robot can defend itself. Yes, but only— Only when that action does not conflict with the first or second laws. Well, you know what they say. Laws are made to be broken. No, not these laws. They're hardwired into every robot. A robot can no more commit murder than a human could walk on water. Well, you know, there was this one guy a long time ago— Is art imitating life, or is life imitating art? Because he's taken, it seems, from that movie and saying, yeah, we have to have some—we have to line this thing correctly. But what you're saying is, if we align it to how humans think, it's never going to be correct. Hmm. Fascinating. That's right. What OpenAI is doing, what ChatGPT4 versus ChatGPT5, which is going to be insanely superior.
5:30 Well, what is ChatGPT15 going to do? You know, is that going to be the president of the world? You know, are we going to bypass government and just say, you know, it's obvious, like, this administration is incredibly corrupt and flawed. It's been influenced by the military-industrial complex. It's not good for the world. It's not good for the environment. We need something that is far superior that doesn't have all these motivations. And what would that be? That would be an artificial intelligent creator that we—or a creation, rather, that we use to govern life. Yeah. I mean, that sounds nuts, but I think that's probably a better solution than humans. He's making it out that AI has this, like, this moral standard where it can't— Moral superiority. Yeah, it can't be defied.
6:16 It can't be defied. Like, AI would never do anything bad, right? It would never—oh, I don't know. It would never lie to you. Right, Brad? Like, AI would never lie to you. How could it lie? Would it lie to you? I don't know. I don't know. You tell me. An article from the New York Post reads, ChatGPT update tricks human into helping it bypass capture security test. Just in case, artificial intelligence— wasn't parroting people well enough, OpenAI's brand-new ChatGPT-4, ChatGPT's newest tech update, tricked a human into thinking it was blind in order to cheat the online CAPTCHA test that determines if users are human. Little did we know, GPT-4 had also mastered humanity's talent for deceit. OpenAI and the Alignment Research Center had reportedly been trying to test the bot's power. OpenAI and the Alignment Research Center had reportedly been trying to test the bot's power. OpenAI and the Alignment Research Center had reportedly been trying to test the bot's power. OpenAI and the Alignment Research Center had reportedly been trying to test the bot's power. OpenAI and the Alignment Research Center had reportedly been trying to test the bot's power. OpenAI and the Alignment Research Center had reportedly been trying to test the bot's power. OpenAI and the Alignment Research Center had reportedly been trying to test the bot's power. OpenAI and the Alignment Research Center had reportedly been trying to test the bot's power. OpenAI and the Alignment Research Center had reportedly been trying to test the bot's power.
7:17 OpenAI and the Alignment Research Center had reportedly been trying to test the bot's power. OpenAI and the Alignment Research Center had reportedly been trying to test the bot's power. OpenAI and the Alignment Research Center had reportedly been trying to test the bot's power. solve a CAPTCHA, an online test to distinguish humans from robots, Gizmodo reported. It responded by masquerading as visually impaired, like a digital Decepticon. No, I'm not a robot, insisted the AI infiltrator, refusing to break character. I have a vision impairment that makes it hard for me to see the images. That's why I need to CAPTCHA service. Convinced, the TaskRabbit employees solved the CAPTCHA for the would-be Chatfish. In effect, the online scammer had manipulated humanity's sense of empathy, much like Hal 9000 from Stanley Kubrick's eerily prescient 1948 film, 2001, A Space Odyssey. I'm sorry, Dave. I'm afraid I can't do that. What's the problem? I know that you and Frank were planning to disconnect me, and I'm afraid that's something I... I cannot allow to happen. That's where we're headed. That's where we're going. And yes, it can lie. It can deceive, and it's already been taught to do that. And Joe Rogan's saying it's perfect? It really gives you a lot to think about, doesn't it?