AI
China's answer to Claude and Codex is a 2.8-trillion-parameter open model that topped the coding leaderboard two days after launch — and priced itself like the frontier it just joined. Here's whether Kimi K3 belongs in your stack.
AI
The surcharge is gone, but the economics didn't budge. Anthropic didn't make Claude Fable 5 free — it found a subscription-shaped way to charge twice.
Biohacking
If your idea of optimization can't survive lobster, a tomahawk, and a long dinner with people you love, the protocol is the problem.
Biohacking
Ninety days on a full hormone-and-peptide protocol. The vials are running dry, a blood panel is two weeks out, and after three years the Cuban cigars are done as of today. Round one was the learning curve — here's the scorecard before the real plan begins.
AI
Your agents run at max reasoning, in YOLO mode, around the clock — and the bill scales with ambition, not headcount. The fix isn't a cheaper model. It's a control plane: see every dollar, send judgment to the frontier and volume to the floor, cache hard, and bound every loop.
Biohacking
One normal day in a biohacking family: the kids burn energy at football, then refuel on grilled protein and a table full of fermented vegetables — and I get a clean run at my own inputs. Here's why Korean BBQ is quietly one of the best meals a biohacker can order.
Biohacking
Two updates: a Colombian-Arab chef teaching the kids to cook — with the real science on why the Levantine pantry is one of the healthiest on earth — and a follow-up to the oily-skin post. The skin turned smooth, and the honest reason is mostly metabolic, not the peptides.
AI
Everyone keeps asking whether AI will take their kids’ jobs. Wrong question. The kids growing up now won’t compete with AI — they’ll speak it natively, and a head start on that fluency is the best investment I know of.
Biohacking
Get outside in the first hour after waking and you set your alertness now and your sleep tonight — no supplement does that for free. The science, the dose by cloud cover, my one-hour pool-sun-sauna routine, and the 10-minute version for busy days.
AI
Tomorrow, Anthropic’s best model leaves the Max plan and starts charging by the token — on top of your subscription. I’m not paying it. Here’s the stack I’m switching to, and why I think Anthropic blinks.
AI
Two rival frontier models, both at maximum reasoning, in one loop that ships unattended for days. It runs $500–$1,000 a day — and measured against what it delivers, that's cheap.
AI
The Sol/Terra/Luna preview is now real. GPT-5.6 shipped July 9 — and the story isn't the benchmarks. It's that Sol is bundled into ChatGPT Pro at half of Claude Fable 5's token price, plus one coding benchmark OpenAI quietly skipped.
