When OpenAI Buys the Chips, Everyone Else Buys the Hype

OpenAI's deal to buy hundreds of thousands of AMD chips worth the computing power of six nuclear reactors, plus a 10% stake option, deepens its dominance and draws major investors like Microsoft and SoftBank. TJ Walker warns the concentration could stifle competition and innovation in AI, especially given the stakes of the technology.
Chapters
OpenAI bets big on AMD and everyone else bets on OpenAI. The AI arms race just took another massive turn. Reuters reports that OpenAI will buy hundreds of thousands of AMD chips, roughly the computing power of six nuclear reactors, as part of a new equity for silicon deal. In return, OpenAI gets the option to acquire up to 10% of AMD. The new set AMD stock soaring more than 35% in pre market trading. The deal cements OpenAI's position as the clear frontrunner in artificial intelligence, and investors are treating it that way. Tech giants like Microsoft, Nvidia and SoftBank have already poured billions into OpenAI's ecosystem, while smaller players scramble to catch up. But such dominance brings new questions. With OpenAI's compute capacity now dwarfing most rivals, regulators are watching for signs of monopoly power in what could soon be the most valuable industry on earth. In short, everybody is betting on OpenAI. And if history is any guide, the front runner often ends up writing the rules for everybody else. What's the real message here? OpenAI is a train that's getting faster and more powerful every day as more and more other big tech companies are saying, this train is leaving the station, leaving, let's hop on, let's invest in them, let's give them money, let's give them a part of our business. It's because they are afraid that in the new world of AI, there's going to be one winner, everyone else is going to be a loser. That may be smart business decision for these companies trying to get a piece of it, but it's not necessarily good for consumers. If the last 25 years has taught us nothing else, it's when one company is too dominant, problems emerge. Let's dig a little bit deeper here and full disclosure, I use OpenAI. OpenAI helped create the graphic you're looking at right now helps me refine and rewrite some of the scripts every day. I do their inexpensive 20amonth version. So I don't want to be a hypocrite here and say how Dare anyone use OpenAI? I use it every day. I do find it's. I don't even know why I started using that as opposed to any, but I did. It certainly got the most attention early and I, I logged onto it, I guess. But now we face a problem where the company is so big it's now given a valuation of half a trillion dollars. It's not a public company, it's the most valuable private company in the us and in some ways, other than a few royalty Funds. It is the most highly valued private company in the world. So with that much wealth comes tremendous power. And at some point, maybe not with the current administration, but with other administrations, with your half a trillion dollars, maybe a trillion dollars in a few years, you start owning the politicians, you have so much money you can buy them all off. Regulatory issues go away or fade away. Quite often antitrust issues, certainly in the current environment, disappear and it leads to tremendous consolidation. Small players are snuffed out. Innovation can be snuffed out. Here's the big problem with AI that's simply different from one from Google snuffing out competition or Facebook snuffing out competition. The problem with AI is they can be the first one to truly get general artificial intelligence. A level of artificial intelligence so smart it's truly surpassed humans ability. And the big fear of course, is when you look at the history of all animals on planet earth, there's never been a case when the smarter animal doesn't rule over the dumber animal. So many people, not just conspiracy theorists and dystopians, many people who are serious AI experts, are worried about the fact that once a general artificial intelligence is created, it is going to be the master of human beings. Maybe it's mastery is benign, maybe it's a benign dictatorship. But would we as humans knowingly hand over our autonomy, our power, our own sovereignty as a species to something else? Because make no mistake about it, artificial intelligence is something else. It's not human. So that's a part of the real concern here about OpenAI getting more and more companies wanting a piece of it. And when you're, when you own a piece of a company, you're less likely to come up with competing ideas. Some of the landscape now that's very fertile in the AI space is for companies trying to really take safety seriously. Because OpenAI is kind of like eh, we used to talk about safety. Safety is nice. But you know what else is nice? Raising another $250 billion. That's what's that seems to be the attitude of Sam Altman and The leadership at OpenAI today is safety is a nice thing to have. Raising more money and getting richer and building our war chest is even more, and growth is more important. And maybe it all works out. Maybe 500 years from now humans are all incredibly happy and writing poetry and doing relaxing hobbies all day long and everyone's happy and AI is doing all the manual work, all the other work and it all works out. Maybe, but maybe not. And that's why I do think we need more AI companies. We need competition for OpenAI now. There is a lot of competition right now. But as many tech analysts predict, we're going to see what happened in the Internet era and in the railroad era, in, in Britain and in the automobile era 100 years ago. New technology happens. Everyone's excited about it, investors all fall in. And then after 10 years or so, it is clear, okay, 98% of these companies are going to fail. Too many railroads were built, they failed. Too many car companies were built, they failed. Too many Internet startups, pet.com failed. And you're left with three. Sometimes in the case of automobile manufacturers in the United States, for a long time, sometimes you're left with one. Like in the case of Google. If we only have one OpenAI company, or see, I'm already doing. I'm conflating AI with OpenAI. That's essentially a win for them. That's what they want. If we only have one AI company, is OpenAI the one we trust the most? I don't hate the company again. I use it every day. But I don't have anything to really give me confidence that they should be the ones potentially controlling the entire future of humanity. I know that sounds like a lot sounds heavy for a Tuesday morning, but I'm not the only one who feels like the stakes are that high.
More from TJ

You're Not Being Judged on Your Work Yet: What New Hires Are Really Evaluated On









