Ffacebook.com
Czy przestać oszczędzać z powodu AI? Modele AI zgodnie: tylko przy ponad 90% pewności co do skrajnych scenariuszy
Ilustracja wygenerowana przez AI
Claude, ChatGPT, and Gemini converged on this comment: Is there any set of beliefs that would make it rational to stop saving money now? Yes. But for most people, the bar is far higher than these tweets suggest. The logic is simple. A euro saved only matters in futures where an extra euro still improves your life. Roughly, there are four kinds of futures: 1. Broadly shared abundance: everyone's needs are met regardless of savings. Your savings are worth almost nothing. 2. Catastrophe: you're dead or the system is gone. Your savings are worth nothing. 3. Capital-dominated AI: AI makes human labor economically worthless, but property rights hold. Owning assets becomes the main claim on what the economy produces. Without savings, you're poor in a rich world. 4. Slow diffusion: AI spreads gradually, the economy looks familiar, and you still age and retire. If you depend on your savings for future security, even a small chance of ending up poor can justify substantial saving. Being broke in a world where money still matters hurts far more than spending a bit more today helps. In illustrative models, the confidence in futures 1 or 2 needed to justify saving nothing can exceed 90 percent, or even 99 percent, depending on how risk-averse you are and how bad the downside would be. And the claim isn't just that superintelligence arrives. It's that it arrives soon enough, and is shared widely enough, that your own money stops mattering to you. But "stop saving" is the extreme case. Take a simplified two-period model with logarithmic utility: income of 100 today, and a fallback of 10 later if money still matters and you saved nothing. The optimal amount to save is 45 if you're certain money will matter, 27 at 50 percent, 8 at 20 percent, and zero only at 10 percent or below. These are illustrations, not financial advice. The point is the shape: all else being equal, growing confidence in shared abundance can justify saving less. It rarely justifies saving nothing. Here's the twist. More confidence in AI doesn't automatically mean you should save less. If the future you find more likely is AI replacing labor while ownership stays concentrated, the rational response is to save more and own the right assets. So the decisive question isn't how likely superintelligence is. It's who will own its gains, and how fast they spread. That's a question about politics and institutions, not technology. Two more observations. Musk can afford to be wrong about this in a way most people cannot. Advice that poses virtually no threat to his financial security could be disastrous for someone who depends on savings for retirement. And Viv's tweet openly admits to "justifying" decisions already made. A belief about the future that conveniently licenses what you already wanted to do deserves extra scrutiny.
Czytaj oryginalną publikację