What I’m up to, Now.

Heard from VC LPs this week: “An OpenAI or Anthropic direct allocation is the Birkin Bag 👜 of investing and the multi-layer SPVs are the knock offs” 🤣🤣🤣

Join millions who have switched to Grok.

Internet -> Websites Cloud -> SaaS Mobile -> Apps AI -> Agents?

End of 10-day China trip; sole objective: learning to gain perspective. Thankfully, @danwwang kick-started that effort by sharing his new book "Breakneck." Perfect lens into modern China with a mirror pointed back on USA. Great writing. Must read.

Why AI is a house of cards: 1. You pay $200 a year for an AI app (like Cursor). 2. Cursor pays OpenAI $500 for API tokens ($300 of which is VC funding). 3. OpenAI pays AWS $1000 for compute ($500 of which is VC funding). 4. $AWS pays $10k for $nvda GPUs. See the problem?

@thomasahle LLMs work well for many cases, but in these cases, they aren't the right tool: - Latency/throughput sensitive - Need layout/positional info - Absolute accuracy and low hallucination risk matter - Edge cases like complex tables, etc. - Want deterministic customization -

For some reasons when consumer companies do it people think it’s more absurd. Just remember - selling something for less than it costs boosts demand. Has to.

Every Prompt is a Product

https://t.co/OH70EI6BRG

For this episode @jeremyau interview @tianglim from Forge Ventures regarding the VCs and startup scene right now especially with the recent scandal and AI gold rush. My biggest take was "Build in Asia, serve the West market!" https://t.co/PReNYXjzC1

There’s a window right now where AI agents will get built for every vertical and domain. The playbook is to go deep on the context engineering required for the vertical or particular space, figure out the right UX that ties into the existing workflows naturally, and connect to

Current AI “alignment” is just a mask Our findings in @WSJ explore the limitations of today’s alignment techniques and what’s needed to get AI right 🧵

If something feels too hard, it just means that the first step isn't small enough.

Here is a ChatGPT summary of the paper: The Illusion of Thinking: Understanding the Strengths and Limitations of Reasoning Models via the Lens of Problem Complexity (Apple ML Research) 🧠 What the paper is about Apple introduces Large Reasoning Models (LRMs)—language models

When the economy is strong, it’s a good time to sell things. When the economy is weak, it’s a good time to build things.

Sad to see this one go

Did a fireside chat yesterday at an AI conference, and then got a bunch of very polished emails summarizing key points of my chat, telling me how insightful I was, and asking to meet. Clearly: audio intake --> transcript --> AI --> email. Interesting to see these little

new one minute blog: the big lie about competition

😂😂😂

2.

The First Four Years