I think what we see with CRWV getting upfront payments from the companies that will benefit from their success, and with that AKAM deal where converts participate in a share price increase, we are starting to give away upside in return for financing. Which makes sense.
I feel like the basic problem with AI financing has been trying to raise capital from credit markets in a traditional way, meaning no real upside and you manage the downside risk thru yield. But AI is lots of upside with lots of downside. It makes no sense to cap the upside.
2. We believe that a majority of this acceleration in capability improvement is likely now driven by AI being used in AI R&D. We aren’t sure of this, though, and did not find consensus about this among Anthropic’s staff. Non-AI factors likely played a large role in late 2025.
An “acceleration” would mean that AI capabilities increase even faster than their historical trend.
Our assessment studied three questions: (1) is acceleration occurring, (2) is it attributable to AI, and (3) are there blockers to near future growth?
https://t.co/ErENKaNEew
@JohnHuber72 Because I think what you must be saying is that CRWV is a proxy for AMZN here. We can learn something about AMZN by watching what happens to CRWV. And I'm not sure I understand what that is.
@JohnHuber72 So just to play devils advocate, let's say CRWV fails as a company because they can't raise funds. Do you think this changes the investment opportunity for AMZN in any way? Is it a negative or positive for AMZN?
I was a thermal engineer in a former life and if I was concerned with thermal expansion on a big wafer, adding dummy work to keep power relatively constant and limit thermal fluctuations might actually be a solution to a problem? As long as I could get rid of the heat. $CBRS 🤷♂️
One neuron out of 1.6 million turns Qwen3-32B into a momentum stock trader that ignores fundamentals. Turn up the L43:12140 neuron and it cares way more about the technical setup than the fundamentals of a stock, the effect survives multiple ways of asking it's opinion.
The most advanced AI adopters are now spending *less* on AI as they advance along the adoption curve.
We can't see the scales on these charts but based on Ramp data among top 1% firms, it's possible Notion has engineered an outright decline in total AI spend.
This is a fundamental reversal of how we previously forecasted adoption: More adoption >> more usage >> more spend. Past a certain point, the most advanced businesses also figure out how to *control* that spend, and with simultaneous price cuts, costs come down again.
I feel like the fact that compute scales linearly with model size but quadratically with sequence length is under the radar. A token is a token but not really.
@VictheDope Nope. Because I have already voted against it on all accounts. But that doesn't mean I can't look objectively at a story that is framed to incite fear and call it that. If your point is that I have to side with poor journalism for a greater good, then I don't believe you.