Writing
Long pieces, in no hurry, written to understand rather than to conclude.
Featured
I dropped the PunditBench LLM predictions into my friends' World Cup pool, scored two ways. On raw scores DeepSeek V4 Flash edged everyone; under a betting-odds system only Grok 4.3 cracked the top five. The lesson isn't who won: the same market finishes 55th played naively and 20th played well, so turning a prediction into a good ticket matters more than the prediction.
Photos from a June week around La Rochelle: Fort Boyard from the sea, the islands of Ré and Oléron, Rochefort and its transporter bridge, and older stone inland at Saintes.
I re-ran the 21 CodinGame code golf puzzles from the Opus post with Claude Fable 5. It wrote 24.7% fewer bytes overall, 10,712 down to 8,065: still above the best human TypeScript, but closing the gap fast.
My static Astro blog took 12 minutes to build and deploy. The culprits: satori-html parsing 320 KB data URIs quadratically, AVIF encodings thrown away with every CI runner, and an upload that re-sent unchanged files. Three fixes, a few lines each.
I pointed Opus 4.8 at six CodinGame optimization puzzles. On the travelling salesman it tied the top of the leaderboard; on the open-ended games it reached a sixth or an eighth of the best score and stopped. The thing it's bad at is the CodinGame squeeze, not the solving.
I ran Opus 4.8 against 21 CodinGame code golf puzzles in TypeScript, from easy to hard. It never came close to first place, and usually didn't even beat the best human TypeScript. Here are the numbers.
Two days at Anthropic's Partner Basecamp: the diagnostic loop, evals as a discipline, the three phases of AI inside a company, and an evening to rewalk London.
Photos from a late-March trip around Narbonne and the Côte Vermeille: Carcassonne, the canal at Narbonne, a rhino at Sigean, and the wind on the coast at Collioure, all off-season and under the sun.
