Don, there is a missing term in your equation. The crunch spread is not: Token Price - GPU rental price. Rather it is: (Token Price * Token Volume) - GPU rental price. More simply: Revenue per GPU - GPU rental price. Continuous throughput improvements for the same exact token mean that volume growth can more than offset token price declines for the exact same model/token i.e. Kimi K3 token throughput per GPU has recently been improving at a 37% annualized rate driven by inference engine optimizations. GPU prices are vertical because: 1) Mix is shifting from higher margin frontier tokens to lower margin open-weight tokens. $10 in Claude revenue requires $1 of inference compute. $10 of of open-weight revenue requires $3 to $7 of inference compute given competition between inference providers leads to lower margins. Open-weight tokens taking share is positive for AI infrastructure demand given lower margins at the model layer. The reverse is also true in that frontier tokens taking share is negative for AI infrastructure demand. Open-weight models have been taking share since early summer. 2) Throughput is steadily improving for each individual token type such that GPU compute becomes more valuable even as true like for like token prices decline.

Last month, we demonstrated high-rate alchemy: turning mercury into real, stable gold, beating CERN’s record by a factor of 10. Last week, we enriched isotopes in the world’s most powerful plasma centrifuge, beating the Soviet record that stood for 40 years. Now, we’re building a fusion reactor in San Francisco to achieve record power density.

The looting in Paris today.. France riots

“Green policies” that make them sound cool at Davos are one of the biggest paper clip objectives btw Second is “rescuing oppressed minorities”

Investor update: I don't often make equity trades (my average hold period is several years), but I wanted to share that I sold all my $VOO and $SPY yesterday morning. Why? The AI trade is the only thing holding up the economy right now, and that trade now has too many ways to fail for it to be safe. Increasing interest rates is new information which surprised everybody, and the second-order impacts of this are not being appropriately accounted for. Here are the points which guided my decision. 1. Inflation stays higher for longer. Interest rates stay higher for longer. Bond prices for data center bonds plummet. Bond holders will need to finance coupon payments with equity sales. Because of all of the self-dealing and trades with equity in the AI trade, this will lead to a fast unwinding in AI stocks. 2. AI spend or use will reach saturation much sooner than we think. Many at the frontier assume usage is going exponential from here. But from where I sit, usage is stagnating amongst early adopters, and companies are being much more critical about spend. Models may be getting smarter but they are not getting cheaper, and ROI is still unclear. But the key reason why this is hidden in plain sight is that no one wants to admit this, because it makes them look stupid. Use might even still increase--just linearly--but that will not be enough to prop up demand. 3. Lab IPOs flop, or get cancelled altogether. This happens, and everything else falls apart. People will lose confidence and the impacts of this will ripply throughout the entire tech ecosystem. 4. RSI becomes real and everything else trends to 0. If a lab creates a superintelligence, not sure how we price everything else in the broader market. For instance, if superintelligence invents a new AI chip and coordinates the operations and starts running on a better GPU than NVIDIA makes, what happens? 5. A recession happens. Interest rates jumping this quickly, combined with high debt loads, are a bad combination at the corporate level. I also think consumer behavior could break with the cost of money being so high. The U.S. runs on credit and high inflation was very unexpected a year ago. I do not think the risk for this is priced in enough. And I don't think it's just because of the war in Iran. For context, I didn't take this decision lightly. I had held some shares of SPY since 2005 (!) that I just sold yesterday. I was proud of that. Those were some of the first index fund purchases I ever made. But I won't let sentimentality guide my investment decisions. Overall, I do not think this the current situation is sustainable. Putting a huge proportion of assets into the S&P 500, of which 40% of holdings is comprised of just 10 companies, all of which are highly correlated and at risk due to the points above, is an irresponsible decision. I still hold some individual stocks which are significantly better positioned against the above risks, like $BRK.B, and long-term zero coupon U.S. treasuries.

The statistics said MAID high an all time high in Canada so I asked how many Canadians are left and they said they just keep getting more from India so I said it sounds like they’re just slowly genociding the entire world and then Carney started crying

The reading comprehension skills in the comments are so low that it’s no wonder they’re looking for educated & skilled people. What part of “high-wage” jobs, which involve bringing in highly skilled & educated immigrants to areas where we lack them, don’t you understand? Go look at the job list & see how many of you qualify.

I’m not asking for much. Just the death penalty for every UI designer who builds a mobile interface that shows you a bunch of buttons and then as soon as you go to press one it rearranges itself, so you do something totally different.