Leonid Zadorozhnykh @ProDoser: "The race nobody can afford..."

Finland

The race nobody can afford to quit The strangest thing about the AI race is that the people building it are often the same people warning us about it. Musk warns about the danger, then opens xAI and builds Terafab. Altman warns about it, then raises the next hundred billion. The team at Anthropic built their entire identity around the risks, and they keep training bigger models anyway, the latest one had to be blocked by government already. Everyone near the front of this race agrees, out loud, that slowing down might be the wiser choice. And not one of them slows down. That is not hypocrisy. It is math. There is a classic puzzle in game theory called the Prisoner’s Dilemma, where two people would both be better off cooperating, yet the logic of the situation pushes each of them to betray, so both end up worse off. If you have not seen it before, I’ll add link in the comments. The reason it matters is that humanity keeps stumbling into this same trap at the largest possible scale. The nuclear arms race was a Prisoner’s Dilemma between superpowers, where both sides kept building far past the point of any sense, because stopping first felt like surrender. The Space Race to the Moon was the same shape, two nations pouring national fortunes into a sprint neither could afford to lose in front of the world. Go back further and you find it in the naval arms race before the First World War, when Britain and Germany bankrupted themselves building battleships nobody wanted but neither could stop building. It was always like this in the past. The AI race has that exact structure. If one American lab pauses, another passes it. If America as a whole pauses, China does not. Nobody can pick the cooperative outcome on their own, because picking it alone is just losing. I will be honest with you. I do not have a clean way out of this, and I do not think humanity has found one yet. You cannot stop competing, because stopping is losing, and the competition itself is racing toward something none of us can clearly picture. But I want to separate two questions, because they are not the same thing. There is what I feel as a citizen, and there is what I see as an investor. Here at eToro I speak as investor. Whatever you believe about the ethics or the endgame, the money has already moved. More than a trillion dollars is locked into AI buildouts, and nobody is writing that off. Chips from $NVDA (NVIDIA Corporation) sit at the center of it. The hyperscalers behind the labs, $MSFT (Microsoft) among them, keep raising their spending rather than cutting it. $META (Meta Platforms Inc) is pouring capex into the same fire. And underneath all of it sits the least glamorous, most essential layer, memory, where $MU (Micron Technology, Inc.) lives, because none of these systems run without it. Then there is the part that used to sound like science fiction. Terafab, the $TSLA (Tesla Motors, Inc.) and SpaceX chip project, is real, aiming at a scale of compute the entire industry cannot currently supply, with most of its output pointed at space. Orbital data centers are no longer a thought experiment either. $SPCX (Space Exploration Technologies Corp) has already filed to launch compute satellites, and it is not alone, since Google and Amazon are studying the same move. A few years ago this was a slide at a conference. Today it reads less like fantasy and more like a roadmap for the second half of this decade. So my outlook here is not built on hype. It rests on one plain observation. This is a race that no serious player can afford to leave, and capital tends to follow that logic rather than fight it. I do not see a realistic mechanism that slows it down in the next few years. Do you see a point where this race actually slows, or do you think we are locked into the game for the rest of the decade? ——— 📊 This is not financial advice, and I am not a financial advisor. Always do your own research. 🤝 If this way of thinking resonates and you want to follow the journey, you can copy my portfolio from a $300 minimum. ⚠️ Past performance is not an indication of future results.

Not investment advice. The author may have financial interests in the mentioned instruments.