Where the AI market is talking

OpenAI's Head of Data Centers Has Left the Company

Perhaps everything is fine and this person and the bunch of other recent high profile OpenAI departures just really hate money.

Rumors that OpenAI recently finished new >10T parameter training run

OpenAI Jalapeño: Better than Nvidia Blackwell

Warning, this is a long comment! (I’m trying to stick to sourced facts here and not overstate what they mean) I went down a rabbit hole after watching Dylan Patel on Dwarkesh today: https://www.youtube.com/watch?v=aV26V1UvkJw I was initially just surprised by how bullish Dylan is on OpenAI/Anthropic and how bearish he is on China, despite Chinese labs getting closer to US SOTA while offering inference at dramatically lower prices. So, I started digging while waiting for various day-job inference calls to return, ha. Dylan says he spent years obsessively posting on hardware forums, moderating hardware subreddits, and running anonymous hardware blogs/videos before SemiAnalysis. But he also says most of that history is now gone, including from the Internet Archive, b…

Run Minecraft in a Windows sandbox for computer use agents

I've played around with Mindcraft for a while. Fair warning that it is quite outdated and spaghetti-coded. Although all of its components required to make it work are somewhat hard to replicate from scratch, so it might still be your best option.

Starbase, LA

That’s just a microcosm of everything. The culture war ate the entire public discourse to the exclusion of practical concerns and even coherent ideas. It’s all vibes and tribalism and bashing the other tribes, while the tribes themselves become increasingly incoherent in terms of what (if anything) they really mean or stand for. I just turned it all off, though it makes me fear for the future. Feels like the early stages of societal collapse.

Anthropic Sees over $30T in Potential Revenue

It is pre-IPO public dance. Today, they offered me shares in Anthropic; the IPO is supposedly scheduled for October.

Anthropic Sees over $30T in Potential Revenue

According to Reuters which refers to this WSJ article [0], $30T is the TAM. Calling it potential revenue is a huge stretch. Not justifying as this number as it is stupidly large, but theoretically if AI replaced all White collar workers in the World it will come to around $30T or so (Gemini said it’s $36T-$37T based on some calculations using global GDP, labor share, white collar share etc) TAM is always stupidly large in decks, disconnected from reality. Every startup has a deck with TAM in billions. [0] https://www.reuters.com/business/media-telecom/anthropic-exp...

Apple introduces M6 and M5 Ultra

Last thing i heard was that large companies are starting to seriously spend on ai infrastructure (think nvidia clusters rather than tokens from openai or anthropic) to fine-tune and run custom models. This might become the driver for memory and chips demand in the near future. But i don’t know, of course, we’ll i guess.

OpenAI restores 5-hour Codex and Work limits for ChatGPT Plus users

The rare instance I see them is when needing to use the actual Facebook phone app, and then I see the ads targeted at me. Basically wall to wall ads for Cursor, Anthropic, and Littlebird.

OpenAI Jalapeño: Better than Nvidia Blackwell

I think they talked about this being general purpose chip but I would think that Anthropic/OpenAI are at the scale now they could bake LLM weights into chips themselves. For example, GPT Sol baked into a custom chip run for $100M that runs 10x as fast and 10x as cheap should pay for itself as long as the chip is useful for long enough. While 2 years ago nothing was useful more than 1 year long, there are many older models in use now (e.g. Haiku 4.5, GPT-OSS 120b), and I expect this trend to continue. I know this is what Taalas was doing (acquired by AMD), here was their demo, https://chatjimmy.ai/ which is based on Llama 3.1 8B. It feels like this should start to happen soon.

Anthropic expected to tell investors it sees over $30T in potential revenue

Reuters should know better. While $30T is probably absurd, Anthropic is claiming that's the total addressable market for AI services, not that they will capture 100% of that market

New Mac Studio with M5 Max and M5 Ultra

There are signs that the money faucet is being turned down. Various investments that were announced have been quietly canceled or reduced in scale. I don't think it'll happen soon, but it seems unlikely to be more than a couple years. If I were a betting man, 12-18 months seems right. The AI companies that can make enough money will survive, the ones running on investor cash and debt, won't. Efficiency is improving, both in hardware and in software and in intelligence density (smaller models can effectively do more of the AI work that needs doing), so I think the pure data center plays will falter. If there isn't some other business attached, they're never going to recoup their investment. Anthropic and OpenAI are buying all the compute they can find right now, but efficiency gains, espec…

OpenAI: GPT 5.6 Sol price reduction (until at least Nov 21)

> The fact that AI models can be so easily distilled and replicated is such a stroke of luck. Why shouldn't it work though? It's just models teaching other models same way humans are.

OpenAI Jalapeño: Better than Nvidia Blackwell

SemiAnalysis’ founder was roommates with Anthropic people, not OpenAI, so he may be slightly (very slightly) more objective here.

Anthropic expected to tell investors it sees over $30T in potential revenue

Anthropic tells staff to work from home due to possible security team strike

At this point if news broke Anthropic’s office was bombed my first thought would be the US government. Didn’t they label the company a national security threat?

Anthropic tells staff to work from home due to possible security team strike

Pretty crazy. The security team denies there ever was a threat of a strike? And generally you do announce strike threats, it's kind of the whole point with strikes that your employer knows of the possibility. Did Anthropic get "warned" about this strike by its own AI models?

Apple introduces M6 and M5 Ultra

Fable would have a much bigger userbase if Anthropic didn’t remove a zero data retention option because muh safeguards

The AI Hater's Manifesto

This is a sincere and vulnerable question but can somebody help me understand what kind of mindset it takes to write something like this and not feel like a total scummy scammer person: > If you liked this piece, you should subscribe to my premium newsletter. It’s $70 a year, $17 a quarter, or $7 a month, and in return you get a weekly newsletter that’s usually anywhere from 10,000 to 18,000 words, including vast, detailed analyses of NVIDIA, Anthropic and OpenAI’s finances, and the AI bubble writ large. It’s so overwhelmingly clear to me that people who do this, get paid. So its clearly effective. I’d love somebody else who’s figured this out to explain it to me.

Anthropic tells staff to work from home due to possible security team strike

You don’t think Anthropic belongs on that list? They’re working as hard as possible to plunder most of human creation and disturb all human life so that they can enrich themselves.