All posts

recent

Superpersuasion will look like bribery

October 3, 2026 │ ai, ai safety

Do not build the LLM torture factory

October 2, 2026 │ steering, model welfare

Human-AI partnerships are for alignment, not capability

September 27, 2026 │ ai

Advice to a beginning software engineer

September 26, 2026 │ tech companies, ai

You should all be asking way more questions

September 25, 2026 │ good engineers, ai

System One models like Jev can train their own replacements

September 20, 2026 │ system one, ai

Grit your teeth and ship it

September 20, 2026 │ shipping, emotional regulation

Two techniques for working with System One models

September 18, 2026 │ ai, system one

Jev means structured output is interesting again

September 16, 2026 │ ai, system one

Tell agents the why, not just the how

September 15, 2026 │ ai, prompting

Slow developer experience will bottleneck fast models

September 14, 2026 │ ai

AI is breaking our proxies for expertise

September 13, 2026 │ ai

Don't build tools for AI agents

September 12, 2026 │ ai

They really do think AI might kill everyone

September 10, 2026 │ ai, apocalypse, ai safety

Why we should anthropomorphize AI agents

September 9, 2026 │ ai, anthropomorphism

sean goedecke

popular

LLMs reward expertise

How I ship projects at big tech companies

Software engineering may no longer be a lifetime career