ExclusiveWorld's FIRST AI Government Official; OpenAI & AMD Partnership

The danger I see here is like it opens up a lot of room for manipulations. In this case, you're saying, oh, we're doing this to fight corruption because the previous real person was corrupt. You would think that, oh, it's an AI system. It's trustworthy. It's not going to get money for its own. It's even harder to detect. Welcome to Silicon Valley Front Row. Another episode. I am Steve Ispas, one of your hosts. And I'm David Lam, and we're in the epicenter of Silicon Valley. So Steve, there's lots to talk about, especially this week. What do you want to start off with? Well, you know, there's big, some big announcements. And one of them was NVIDIA investing a hundred billion dollars into open AI. And that's with the B billion, not million. And then days later was AMD investing into open AI an undisclosed amount. It says it's in the billions as well. So to me, this is like, wow, two of these major companies, their competitors, they're both investing a lot of money into open AI to develop further their project. And I guess it's mostly about infrastructure. Now, you kind of looked into infrastructure projects and, you know, what did you find out? What do they need all this money for? Yeah. So we do know that open AI is partnering with these two companies. And with NVIDIA has like over 90% of the market share of these GPUs that supply AI. So it seems like open AI has these big ideas to advance the ecosystem of artificial intelligence and this next generation models, but they need a lot of power to keep it running. And then, you know, with power, there's electricity and water. And along with that is like the land that needs to be used. So we got to think about maybe what would happen to the land and power grid. Yeah. Well, I think it often comes back to power that, you know, many people say that we need a lot of power for AI to function. And I've heard that we are now pretty much maxed out with power and we'll need like 10 times more power even in the next few years to develop all this system. So how is this power going to come about? We also hear a lot of the water that's being needed to cool down the systems. And we're not sure exactly how much of this water is being recycled, how much is able to be reused, but it sounds like a lot of water needs to be used. Now, the announcement right behind you over there about AMD investment into open AI, I think they mentioned the generation in a few years is going to reach six gigawatts of power, continuous power needed to run this data center. So what's the equivalent of that? How do we put that into perspective? So we had to look that up, you know, because it didn't mean much. I mean, six, you know, I know it's a lot of power, but we didn't know what it was. So apparently it's the equivalent of 4.5 million homes. Or a better comparison is the entire energy that New York City uses at one time. That's crazy. So that's probably a significant percentage of the whole country. And I think what's important to note here is that that's AMD's chips that will need that much power in the open AI system. But again, AMD is probably less than 10% of the chips that are being used by open AI and others for AI. The most are coming from NVIDIA. So the number is probably a 10 or 20X would be larger than what we see there. So like 10 or 20 New York City. That's wow. And so in the first year, like 2026, apparently the first gigawatt would be supplied from AMD. That's what it seems like. Yeah, that's what they're saying. And it's pretty crazy because in the first day, the stock value of AMD went up over 20% from like the news of this. Yes. Yes. There's definitely excitement about it. Well, we'll talk more next week about this. We have someone coming to our program, Jim, who's gonna give us more insight that he has about the energy being used and the water needed to run this system. Right. Yeah, I'm excited for that because he also said he studied oceanography. Yes. And also an entrepreneur. So stay tuned for that. Yeah. So I guess we can move on. So the next topic is about Diella. Right. So apparently she's the first AI government official that was appointed by Albania. What are your thoughts on that? Okay, and we're gonna take a quick break. Is that her calling? Maybe we can actually leave. Yeah, no, that's probably Jensen, but I'll call him back later. So this was really interesting. And the first thing I saw this, I really thought it's a joke. And even now it's a little bit hard to understand it, how it actually works. But apparently, this is an actual government official in Albania. And it's an AI. It's a bot. It's not a real person. It's, you know, whatever you want to call it, you know, artificial intelligence being that's a government official. And what's also interesting is that Albania is by far, as I grew up in Romania, right, which is border, you know, very close, not border, but very close to Albania. We always knew growing up that Albania is the poorest country in Europe by far, by far. They're not yet even in the EU. They want to get accepted in the EU. They're trying to work towards that. So this is really interesting that they're the first to kind of get into this. Yeah, they're the first to get into this. And I think there was a lot of pushback, but I still don't quite understand how this will work. Like, you know, apparently, she performed some functions. But the thing is that now you use chat GPT, you ask a question, and you're going to decide whether you, you know, you answer that question, or if you believe that question or not, right. But in her case, she's going to tell you what to do, he's going to say, we're going to invest, we're going to buy this, just charge a procurement, we're going to buy this many things, or we're going to do this. And, you know, basically, that's the sale of the government based on a bond. And now who programs her who, who does all this work behind the scenes, he looks from the article that Bill Gates and Tony Blair are, or Microsoft, Microsoft is is partnering with with the government of Albania, they're backing the the AI appointed minister. Yeah, yeah. So it's very interesting. Again, I thought it's a joke. But to me, the danger I see here is like, it opens up a lot of room for manipulations. I think, in this case, you're saying, Oh, we're doing this to fight corruption, because the previous real person was corrupt, corrupt official in choosing what policies or which contracts exactly. She's, I think this person is judging procurement. So she buys things, right. So there's so much room for corruption there, right? Because you favor some vendor versus another one. So AI wouldn't necessarily be free from hacks. That's exactly my point. Like you would think that, oh, it's an AI system. It's trustworthy, it's not gonna get money for its own. But whoever programs it, and however it's being programmed, however algorithm model is using, it's even harder to detect if there's an issue there, right? Because at least the corrupt person, you see that he has money in his bank, right? Or his family drives brand new cars or something. But in this case, you don't know how the money is being channeled or what happens. Yeah. And you know, what also I read, so apparently, Diella, that's the name of the AI model. She first started as like an assistant bot that helps people with their documents or something, but then got promoted to the government minister. And I guess when that wasn't introduced, and the other officials were, I guess they weren't happy. It was backlash when that happened. And something I also want to like ask is, I guess to the viewers, you know, you could put it in the comments below. How would you feel if a city mayor or government official in your city was an AI? And if that's the case, at what level of information would you be comfortable with it having to process there? Because AI needs data to look through things and to analyze. Well, one more thing to add, David, is that in this case, this bot or whatever you want to call it, is the actual image of an actual actress from Albania that is being used to portray this government official. So the creation itself, it's not an AI creation. It's an actual image of an actress. But we'll talk in a future episode about Tilly, which is actually an AI creation by itself, which doesn't exist as a person who's now, you know... Actress in Hollywood. Exactly. Yeah. So we'll talk about that soon. Another thing, I mean, to ask the audience, how do you perceive this AI development race? You know, I found that there's a survey that said that they surveyed U.S. adults, and over 57% of the surveys said that they see there are high risks for AI. And most of it is because AI could erode human abilities and connections. That's what was commonly said. I think those are probably the main concerns, the abilities and the communications. And of course, from the other side, they probably talk about the efficiency and... And what was the other one? Yeah, I think efficiency is the main... The main thing, becomes more efficient. At what cost? That's a different story, I guess. All right. That's it for this episode. Thank you for watching and hope you join us next time.