The Stock Doc

Three weeks ago there were rumors that one of the labs had completed its largest ever successful training run, and that the model that emerged from it performed far above both internal expectations and what people assumed the scaling laws would predict. At the time these were only rumors, and no lab was attached to them. But in light of what we now know about Mythos, they look more credible, and the lab was probably Anthropic. Around the same time there were also rumors that one of the frontier labs had made an architectural breakthrough. If you are in enough group chats, you hear claims like this constantly, and most turn out to be nothing. But if Anthropic found that training above a certain scale, or in a certain way at that scale, produces capabilities that sit far above the prior trendline, then that is an architectural breakthrough. I think the leaked blog post was real, but still a draft. Mythos and Capybara were both candidate names for the new tier, though Mythos may now have enough mindshare that they end up keeping it. The specific rumor in early March was that the run produced a model roughly twice as performant as expected. That remains unconfirmed. What is confirmed is that Anthropic told Fortune the new model is a 'step change,' a sudden 2x would certainly fit the definition. We will find out in April how much of this is true. My own view is that the broad shape of this is correct even if some of the numbers are wrong. And if it is substantially accurate, then it also casts OpenAI's recent restructuring in a new light. If very large training runs are about to become essential to staying in the game, then a lot of their recent decisions, like dropping Sora, make even more sense strategically. For the public, this would mean the best models in the world are about to become much more expensive to serve, and therefore much more expensive to use. That will put pressure on rate limits, pricing, and subscription plans that are already subsidized to some unknown degree. Instead of becoming too cheap to meter, frontier intelligence may be about to become too expensive for most of humanity to afford. Second-order effects; compute, memory, and energy are about to become much more important than they already are. In the blog they describe the new model as not just an improvement, but having 'dramatically higher scores' than Opus 4.6 in coding and reasoning, and as being 'far ahead' of any other current models. If this is the new reality, then scale is about to become king in a whole new way. It would also mean, as usual, that Jensen wins again.

I wish there were a filter to screen all the AI-written twitter replies. Trolls were the problem before, now the “it’s not x, it’s y” bots make it unreadable.

Your DevSecOps platform shouldn't have a parent company that competes with you on AI strategy, cloud spend, or customer attention. That's not theoretical. It's structural.

BREAKING: Thoma Bravo just released their LP meeting slides. The world's largest software PE firm thinks the market has it completely wrong on software right now. Public markets are panic-selling software based on AI fear. Here's what they're seeing:

Medical billing is broken. We built Coding Intelligence™ to fix this. Automatically generate CPT codes, E/M levels with MDM rationale and ICD-10 diagnoses from your documentation and the latest clinical guidelines. Live now in Visits for verified U.S. clinicians.

Google's TurboQuant... And it's effect on $SNDK, $MU, SK Hynix, and others: What it does: -> 6x reduction in KV cache memory footprint -> 8x Speedup on H100 GPUs It's a compression algorithm. Now... Will it beat down memory? -> Prob not. Implications might be bullish for $ARM and others though where you can run AI locally, rather than DRAM heavy DCs. However: ->This is basically DeepSeek round 3. You can make algorithms more efficient. But that doesn't replace either memory or GPUs. -> It could structurally (and slightly) reduce DRAM demand. -> think it's only been tested on small models so far like Gemma, Mistral, and Llama-3.1 (and paper's been out for a year) Also, markets conflated DRAM with NAND... this algo compresses the KV cache (DRAM). Doesn't do anything to NAND storage? Regardless: Algorithms will always get more efficient. People keep saying Jevons Paradox, which is true since this just scales use cases. Main thing to look out for is hyperscaper capex projections, not Google Algorithms that made things more efficient. Feels more like a narrative headwind than anything material to earnings.

Jevons paradox: Efficiency gains (e.g., better coal use in 1865) increase total consumption as lower costs spur more demand. TurboQuant cuts LLM KV cache memory 6x+ with 8x speed, zero accuracy loss. Short-term: less memory per inference. Long-term: AI adoption explodes—bigger models, more apps, edge use. Total compute/memory demand surges. $MU (DRAM/HBM) and $SNDK (NAND) benefit big. Market's bearish take is shortsighted.