Posted inBlog
The Coordination Tax Most Forecasts Miss
For a while now, there has been a nagging sense that a lot of what gets called "knowledge work" is actually something else entirely. It's not thinking or deciding or…
Posted inBlog
Cheaper Execution Should Make You More Ambitious, Not Leaner
A year ago, the dominant AI conversation in most leadership teams was about efficiency: fewer handoffs, faster outputs, the same work done with smaller headcounts. That framing made sense when…
Posted inBlog
You Built AI Memory. Here Are Six Extensions That Compound.
The question that kept coming in after the piece on building an AI memory system was some version of: “OK, I set it up. Now what do I actually put…
Posted inBlog
Chaos as Signal: What the Moltbot Saga Reveals About Personal AI
The Moltbot rugpull — $16 million in a crypto token named after a renaming accident, evaporated in hours — reads like noise at first. Speculative crypto tokens attach to anything…
Posted inBlog
The Agent Economy Is Already Here. Nobody’s Governing It.
Part 2 of the OpenClaw story was expected to bring more security coverage — more CVEs, more enterprise hand-wringing, maybe a regulatory response. What emerged instead was something stranger and,…
Posted inBlog
The Capability Overhang Is Real: Five Things That Shipped While Most of Us Weren’t Looking
There's a particular kind of cognitive dissonance showing up in conversations about AI lately. Someone will say "these models aren't really that useful yet" in the same week that someone…
Posted inBlog
Every Correction Is Either Lost or Compounded
For a long time, the assumption was that the problem was the prompts. If the output felt generic, the solution seemed to be asking better questions. So: reading the frameworks,…
Posted inNews
The Bet Hidden in Anthropic’s Model Spec
For a long time, most of us working in AI thought about alignment the way we thought about software compliance: enumerate the rules, build the guardrails, check the boxes. The…
Posted inBlog
Why Play Is Producing Better Software Than Planning
For most of my career, building software felt like planning an expedition. You mapped the route before you left. You wrote specs, estimated timelines, reviewed the plan with colleagues. The…
Posted inBlog
The Leverage Is in the Specification
In one recent experiment, an AI agent received a 200-line markdown file and ran with it. Sixteen hours later, a working feature existed where there had been none. The markdown…