← Archive

Mighty Brief

Wednesday, May 13, 2026

Start Here

  1. Linear says 60%+ of plans now include agent users — The more important idea is not “agents everywhere.” It is Linear trying to become the shared system where third-party and homegrown agents work alongside teams. That is the right wedge if organizational context beats solo copilots.
  2. Richard Socher’s Recursive Superintelligence reportedly raised $650M+ at a $4B valuation — Big capital is still chasing frontier narrative density. Worth watching less for Socher specifically and more for what it says about how much market appetite still exists for first-principles AI platform bets.
  3. Former Alibaba/Qwen researcher Junyang Lin is reportedly raising for a new AI lab at a ~$2B valuation — Another sign that top model talent is becoming fundable as its own asset class, especially in China where labs keep spinning out around known technical leaders.
  4. Anthropic is reportedly in talks to buy a startup that powers AI agents — If true, it reinforces the stack shift: model labs do not just want API demand, they want control points in the orchestration layer too.
  5. The Strait of Hormuz blockade is now being framed as a chip-supply problem, not just an oil story — Good reminder that geopolitical shocks now hit AI infrastructure through weird second-order bottlenecks, not only through headline energy prices.

BM

Unread scan for dan.wick@blankmetal.ai failed again this morning. gog gmail returned invalid_grant / invalid_rapt. The Workspace fallback also lacked usable credentials. No clean inbox-based BM read, so don’t fake one.
Memory still says the active BM board is Rally, HR&A, Tungsten, Great Hill, Sentinel, Orbis, Care Providers of Minnesota, and Edmunds Govtech. That is a lot of parallel surface area for a shop that wins on crisp sequencing.
Core BM tension still feels right: brownfield client work creates revenue and relationships, but it can also blur the sharper AI-native thesis if every week turns into custom enablement instead of reusable leverage.

Munger Observer

Opportunity Cost: the task list is still carrying stale March priorities. Old obligations keep taxing attention even after they stop deserving it.
Margin of Safety: the Minnebar reference in the latest note is stale, but the principle is solid. Protect slack early so one surprise does not colonize the whole day.

Personal Thought

This feels like a cleanup-and-positioning morning. The outside signal is mostly capital chasing agents, infra, and abstraction layers. Inside the house, the better move is probably closing one stale loop so the next serious build thread gets full bandwidth.

Quick Scan