Blog🏜️
Thoughts & (LLM) Threads
Aug 2026Karp's Meaningful ProblemsKarp's case that elite engineering is wasted on consumer apps when it should target defense, energy, and biology — separated from his agenda, and translated for the solo builder.READ →Jul 2026The Geography of MillisecondsHow compression of latency is expanding global compute markets, and why the solo dev who can tolerate 80ms has access to prices the hyperscale tenant will never see.READ →Jul 2026Before You ThinkA conversation about comprehension, consciousness, and the moment machines started understanding faster than humans. From the N400 window to the prefill phase.READ →Jul 2026The Server Still WhispersA two-part case for owning your stack: the cultural history of infrastructure from toggle switches to complexity rent, then the practical playbook for running it yourself on a $10/mo Hetzner box.READ →Jun 2026Why your AI agent needs a second opinionThe single biggest quality improvement I've found isn't better prompting or bigger models — it's having one AI agent review another's work.READ →Jun 2026Why the traveler comes home youngerA ground-up visual explainer of time dilation, the light clock, the twin paradox, and why GPS depends on relativity every second of every day.READ →