[01]
Conceptual integrity and counting lines of code
Examines whether AI coding agents erode conceptual integrity even as they increase lines of code shipped.
// daily-digest · thu 20 aug 2026 · morning refresh · last 72h focus
Ten items on the Anthropic-OpenAI revenue flip, Stripe's OpenRouter logic, and AI coding practice, plus a rationalist detour and a wildcard on traveling brain waves.
— refreshed for you, in 10 items.
Examines whether AI coding agents erode conceptual integrity even as they increase lines of code shipped.
Anthropic posted $11.6 billion in Q2 revenue, surpassing OpenAI's $6.7 billion for the first time.
OpenAI's CFO told staff the company plans to go public by 2027, sooner if growth continues.
Analyzes Stripe's OpenRouter acquisition as a bet on aggregating AI model routing into payments infrastructure.
Argues Stripe's OpenRouter deal is really about controlling AI expense management and capital flows.
OpenAI previewed Private Safety Processing, an automated misuse-monitoring system that discards customer data after review.
OpenAI revoked outside researchers' access to its Trusted Access for Cyber program, citing a technical error.
Scott Alexander explains why he skips the Substack theology debate, questioning "powerful hidden entity" arguments.
Lenny Rachitsky launched a curated job board for product, engineering, design, and growth roles.
Researchers argue traveling brain waves act as a computational engine driving perception and prediction, not noise.