Claim◆Video · 17:17 · 2m
The business scores an 8 out of 30 — it doesn't make that much money for the work involved, can't run without Mark, and has a weak moat because anyone with enough capital could do it.
Sam (host) · My First Million
PredictionArticle
To avoid decompressing and recompressing the entire array on every edit, the history should be split across multiple rows capped at 128 revisions or 3MB of uncompressed JSON each.
A scalability refinement: chunking the compressed history into multiple rows bounds the cost of each subsequent edit. ✦ AI generated
The author (on GPT-5.6 Sol Pro output) · Simon Willison's Weblog · 2026-08-09 · original ↗
To avoid the overhead of decompressing and recompressing the entire array on every edit Sol suggested breaking the history up into multiple rows, with each one containing a maximum of either 128 revisions or 3MB of uncompressed JSON.
Read full article ↗excerpt · fair-use quotation
Related moments
MechanismVideo · 17:08 · 2m
Naively chunking documents by a fixed number of characters loses important context, whereas semantically chunking legal documents along their natural structural boundaries (sections, subsections, bullets) while retaining hierarchy preserves accuracy that a naive approach would leave on the table.
Alex Boucott · The TWIML AI Podcast
MechanismVideo · 29:50 · 2m
State space models rely on linear dynamics because current methods cannot properly scale nonlinear systems into parallelizable tensor computation.
Ramin Hasani · The Cognitive Revolution