Pinned Tweet
Cleaning out my Dropbox and found some @twilio diagrams I made almost exactly 10 years ago.
Every IC software engineer is becoming a manager (of coding agents). The work is basically identical to management of human ICs. Trying to coordinate across projects. Setting down higher-leverage goals. Asking questions about what's been implemented rather than implementing.
Feels like a reasonable time for my periodic reminder that you don't actually have to allow arbitrary JavaScript to run in your browser as you click around the web. You can just allow-list the sites you trust. nvd.nist.gov/vuln/detail/cve…
Am I the last person to realize that when you zoom out far enough, everything is just change management?
The best user interface no user interface. If you can, just do the thing the user wants without making them think.
Andrew Benton retweeted
Introducing @PoeticHQ: a new AI system that executes complex multi-hour tasks with 99%+ accuracy and 10x fewer tokens than agents.
We raised $50M at $500M from Kleiner Perkins, Founders Fund, First Harmonic, and Genius Ventures to build AI that does complex work inside Fortune 500 companies without hallucination.
While code is too brittle, agents are too unpredictable. The work that runs the global economy - anti-money laundering, fraud investigations, underwriting - needs extreme accuracy.
So we built a new kind of software that pairs the flexibility of AI with the predictability of code.
When the world stays the same, Poetic runs fixed code: fast, cheap, identical every time. When the world changes, Poetic uses AI to regenerate its approach and find its way back to the objective.
In one year, we went from zero to an eight-figure run rate as a team of four.
Since then, we’ve scaled the team and executed the highest-stakes processes at AIG, SoFi, and Chime. At SoFi, a large US bank, Poetic reached 99%+ quality on fraud investigations in five weeks.
Could build a pretty interesting proprietary data product just from wandering around sidewalks in SOMA and the Mission recording the various walk-and-talk one-on-one conversations happening in public every day.
Strong Zenefits vibes
A detailed and brutal look at the tactics of buzzy AI compliance startup Delve
"Delve built a machine designed to make clients complicit without their knowledge, to manufacture plausible deniability while producing exactly the opposite."
substack.com/home/post/p-191…
Haven't worked on this in awhile, but just in case teams might find it useful to automate local dev certs I made a small private certificate authority as a template GitHub repo github.com/andrewmbenton/ca
So good!
"Agentic coders are learning the hard way that you can't escape the "narrow interfaces" (read: code) that engineering labor requires; you can only transmute that labor into something superficially different which still demands the same precision."
New blog post: "A sufficiently detailed spec is code"
I wrote this because I was tired of people claiming that the future of agentic coding is thoughtful specification work. As I show in the post, the reality devolves into slop pseudocode
haskellforall.com/2026/03/a-…
