SAM ALTMAN:
00:01:01
I am very nervous that we have an impending, a significant impending fraud crisis because of this. You know, there's obviously some reports now of these ransom attacks, where people have the voice of your kid or your parent, and they make this urgent call. That is going to get so compelling. All right, so you hear that. You know what he had to say, and it strikes me that this is the guy who brought AI to the forefront. And now, he's telling the people who arguably have their hands on our purse strings to be super, super careful. Your reaction to that balance that needs to be struck in our country? I mean, I think there are legitimate concerns around the AI tech is getting better so much faster. They can mimic voices very easily, from the president to your grandmother or whatever. So I think there are legitimate concerns there. That said, a lot of this stuff is very self-serving to a lot of these companies when they're saying, we need free reign to do what we want. We need to be China. We need to sort of protect this, and any sort of regulation will rein us in a way that could be harmful to Americans. So I do actually agree. There's a balance they need to strikeout, figure out how to, like, prevent harms here. But a lot of what President Trump is putting out there today in this proposed executive order is essentially going to give them a lot more runway to do what they want with the products that they're building. More runway? I think so. I think-- Like, more guardrails? That's right. So under President Biden, it was a lot more, let's say, balanced, a little more restrictive for these companies, companies, like Meta, Google, OpenAI. They really didn't like it. They were chafing at what they thought would be harms to their growth, and I think under President Trump, it's going to be-- they get more tax breaks to build the data centers they need. They get fewer sort of reining in a lot of the safety stuff. So despite what you hear from a lot of the CEOs on Capitol Hill, I think safety is taking a bit more of a back seat. Wow. So the win at all costs sort of accusation against the president, you see, as having a lot more teeth? No, 100%. And look, there are legitimate sort of issues around the race with China. China is developing what are called open source models that, basically, have no guardrails around AI compared to, say, ChatGPT, and American companies are really worried that we're going to lose that race with China. So I think that's another consideration that President Trump is saying, if you want us to stay ahead in tech, we need to push the boundaries here. OK, so that's us having this conversation up here. But for everybody else, students, teachers, people who are concerned about their 401(k)s, their bank accounts, where's our power? So this is the great debate that's going on right now. There's a concern among technologists, but really, among the American labor force that, if you're not on board with these apps, if you're not learning how to use AI and prompts properly, that you're going to be left behind. And a lot think about Obama era sort of bringing people into the information workforce. I think there's a legitimate concern there. I think, like, this is how companies are preparing the future of using AI inside of them, but there's a lot of ways that it can go wrong. There's things called hallucinations, which essentially make up information to give you, because a lot of the technologists don't know how these models work entirely. So balance, I think, is really key here. Yeah, Artificial Intelligence, not real intelligence, and the way I define intelligence is a balance between head and heart, a balance between ideologies, left, right. The president calls it woke ideologies, and he's wanting to make sure that any "woke" liberal ideologies are not influencing the messages that we get back from AI. But that can be a double edged sword. Say, Republicans, conservatives are voted out, and Democrats and liberals come in. They could pull the levers as well. What are the risks from that standpoint to making sure that things are balanced when we ask a question on so-called ChatGPT or any other piece of information? It's a great point, and in this executive order we're seeing this morning, essentially being told, no mentions of DEI, no mentions of climate change or environmental dangers, and those are politicized topics, right? Taking a stance by excising that entirely is a political stance. And I think, if you-- when we go to the next administration, whoever that may be, there's going to be this dance of what do we define as non-biased in these systems. And for the past 10, 20 years of development, just determining what the systems are allowed to talk about has been a real strong area of study among technologists. And I think, if this executive order is carried out by the companies and this stuff is-- whether it's diversity and inclusion, or whether it's in the environment, if that's really excised from how the models answer people, that's a stance. And I'm very curious, when the next Democratic administration comes in, what their sort of counterbalance to that will look like. It's a conversation that is happening almost so fast, that it's hard to keep control of, at least in our own minds. In a world that is ever more artificial, that which is real is even more important. It's nice to talk to you. Thank you so much. A real person on set, on the virtual reality set. Look, I'm touching him. He's for real. Thank you so much. We hope to have you back here on CBS News. Thank you. Thanks for having me.
