Invariant

Brave Browser stats MAU: 120.93M (+2.9% normalized to 31 days) HWM DAU: 51.16M (+3.3%) 28-day Avg DAU: 49.61M (+2.4%) Avg DAU:MAU: 0.41 (0.00) Daily user-minutes: 6.81B (+2.5%) (Why 28-day Avg DAU is the best DAU metric: it's weekend-day + days-in-month invariant.) 2/4

Have you read any Knuth or Dijkstra? You can isolate your e2e test pipeline so that it asserts a code change didn't violate any invariant. I use playwright for this. The difficulty is implementing this in an orderly fashion while still shipping code. To use Claude in production you need something li

the good news here is there are cosmologies which are highly legible to AIs pluralistic relational and scale invariant - apply well to ecosystems, companies, humans and models the "bad" news is that they are incompatible with extractor-capitalism and many will hate the change https://x.com/tszzl/

Brave Search stats QPM: 2.32B (+3.0% normalized to 31 days) Annualized (QPM normalized / 31 * 365): 27.37B (+3.0%) QPM non-Brave: 7.74% of total QPM QPD Peak 5/31: 76.76M (+2.1%) QPD 7-day average: 75.60M (+2.9%) QPD 28-day average: 75.12M (+3.0%) 3/4

Codex can 1 shot invariant tests for a 6k SLOC codebase in 8h with a /goal that referencing OOS material from @getreconxyz and @trailofbits I’ll share a skill soon

sui_v1.72.0_1778216436_ci: [address balances] Invariant checks for transaction effects (#26541)