CMUX IS COMING TO KOREA! ����🇷🇰🇷🇰🇷
I'll be in Seoul from 4/23-26
Hosting cmux x AIM Intelligence Hackathon in Seoul
~ April 26th 8am 8pm
~ Free cmux KR MERCH
No IDEs. Only CLIs, AI coding agents, and whatever you want to build. AI safety, technical, and business tracks. $40k+ in cloud credits, Cash Prize on the line.
🥇대상 — 3,000,000원
🥈Developer Tooling — 1,000,000원
🥈Business & Applications — 1,000,000원
🥈AI Safety & Security — 1,000,000원
🎓학생 특별상 — $5,000 Azure 크레딧
https://t.co/NHKEiG4NJ2
Introducing Pace Summer Fellowship.
For compulsive truth seekers who care about doing great work in a contained period of time.
4 weeks. $10k. Two fellows. One publishable research artifact.
@johnloeber The sync in Rome is even wilder than it sounds — lots of research showing people's heart rates synchronize when sharing an experience. 250,000 Romans weren’t just watching the same event, they were physiologically locked in.
Architecture as broadcast infrastructure.
@george__mack The corollary: once a recruiter can describe the role, the returns compress fast. The naming itself attracts competition. This played out with "data scientist" — wordless in 2012, commoditized by now.
@c_asawa Comp is the one thing you can measure on day one. Mission and culture only become measurable after you've lived them.
The irony is you need experience to value the things that require experience to evaluate.
Agree with judgment mattering, and it doesn't just move up, but it gets redistributed too. When an IC can ship what used to take a team of 10, they (should be) making CEO-like resource allocation calls daily. As AI agents improve, everyone's judgment and systems thinking matters more not just the C-suite's.
If you read this far, I wrote longer field notes on Deep Work — including Cal's shutdown ritual and why you need to schedule your leisure time:
https://t.co/xaYYKX5oeF
Are you your own worst boss?
If you managed employees the way you manage yourself, your company would go under.
That's the uncomfortable truth at the heart of Cal Newport's Deep Work — and why most "productivity advice" completely misses the point.
Annie Dillard:
"How we spend our days is, of course, how we spend our lives. A schedule defends from chaos and whim. It is a net for catching days."
What you do today is important — you're exchanging a day of your life for it.
The rest of the internet is heading somewhere similar, just driven by engagement optimization instead of censorship. Algorithmic convergence is producing the same flattening effect — dramatic narratives outcompete analysis because they engage. China is just further along the curve.
I’m certainly not the first to say this, but the answer is likely not "make AI better at GUIs", but more like "stop routing AI through GUIs." Claude Code works precisely because it doesn't click buttons. More systems will be designed for agentic v. human consumption.
(I’m sure that Excel/Powerpoint will be around for quite a bit longer.)