Reading Notes
Links, short summaries, and notes on what is worth noticing across AI, operations, strategy, and company-building.
As AI lowers software development costs, user research shifts from a background support role to the primary driver of product decisions. User researchers excel at interpreting messy qualitative feedback that automated tools miss.
Meta Superintelligence Labs found that adding inference compute stalls without metacognitive control. To fix this, researchers built an agentic meta-reasoning harness that raises software reconstruction pass rates to 71.
Hex cut chart-editing errors from 21% to 3% by combining an 1,800-case evaluation suite with automated hill-climbing. Hex built its Quick Edits feature by having coding agents test changes against 1,800 test fixtures.
A technical breakdown of how OpenAI serves streaming generative interfaces through sandboxed worker compilation rather than running raw model code in the browser.
Explains why defensible vertical AI products require ongoing harness engineering instead of static wrappers.
GitHub is redesigning its core Git architecture to support agent fleets that push millions of commits every day, decoupling compute from durable storage to increase write throughput by up to 35 times.
See how a persistent Claude Code agent caught and contained an active remote code execution attack on an always-on Mac Mini.
OpenAI's head of ChatGPT and Codex explains why personal agents will soon drive most web traffic and why rigid workflow graphs are becoming obsolete.