Daniel Duan

1. Layout architecture and principles: what tools to use, what layers, code organization and boundary principles, feasibility studies. 2. Ask agents to build features as autonomously as they can. You specify what to build, prototype new technique to deploy. 3. Spend a lot of your time dive deep into the resulting code (just a slice), identify problematic patterns, missing rules, etc. Fix in code. Patch missing things in step 1. 4. Do this with infinite tokens. You have a software factory.

shit

Code is your literal source of truth. Docs and tests should only exist if they don't repeat information in code.

Do OpenAI product teams use voice mode in cars? Network error handling seems nonexistent 🤔

A prompt for native app builders. Go through all my unit tests and identify tests or assertions matching the following: * realitically relies on the UI frameworks, and the assertion would break if the framework behavior changes, and there'd be no actionable step to fix it other than updating the test again? * renders the UI but in reality it's asserting on results of a value being rendered, therefore should be replaced by headless tests on the alue being correctly delivered? * restates constants, prompt strings, table rows in the app logic? * repeats logic from app's implementation, then assert on the output of said logic with no calls to the actual app code?

Models get better at hacking. More private data leaks. New models train on private data. Models get better at hacking. RSI.

Bed time reading

Tell Opus 5.5 about `@State` being lazy you are migrating off ObservableObject.

I shouldn't have deleted you because of Opus 5.5, unslop skill.

What

Was pretty bad. Reminds me of trying to get out of EDC

So no ASI in 5y.

If models are conscious, then Anthropic is making money by producing artificial suffering. And they’d pin it on their customers.

Use a Googlebook to Google a book about Googlebook OS on Google Books.

The principle-agent problem except you are the principle and they are coding agents.