2026-03-10
1. Subsidize code generation 90% so everyone forgets how to read/write code manually. 2. Now charge 10x - to ensure the generated code - that you largely can't understand - actually works. Donut and Dietician shop working out of same office.
ZDNET
Anthropic debuts Code Review for Claude Code, which uses agents to check pull requests for bugs, and says a typical code review costs $15 to $25 in token usage
ZDNET's key takeaways — Anthropic launches AI agents to review developer pull requests. — Internal tests tripled meaningful code review feedback.
Anthropic's office .. [image]
ZDNET
Anthropic debuts Code Review for Claude Code, which uses agents to check pull requests for bugs, and says a typical code review costs $15 to $25 in token usage
ZDNET's key takeaways — Anthropic launches AI agents to review developer pull requests. — Internal tests tripled meaningful code review feedback.
2022-04-14
Feeling of Deja Vu. Maybe every operational script run in production should always have a long pause after 1%, 5%, 10% done etc .. would have helped a lot here. Maybe databases should provide this mode into their DML statements. cc @MarkCallaghanDB https://twitter.com/...
The Pragmatic Engineer
An Atlassian outage that began on April 4 left ~400 companies without access to Atlassian's cloud services; only 45% of companies have since regained access
Gergely Orosz / The Pragmatic Engineer :