today we look at fresh tools for developers, from a new coding agent to a guide for the claude api. research highlights include a smoother way to train neural nets and a method to measure lag importance in forecasts. we also note a sandwich chain's ipo filing that shows ai marketing has jumped the shark.
- llm-coding-agent 0.1a0 released - simon willison ships a new coding agent built on his llm library with tools for file editing, command execution, and search.
- getting started with the claude api in python - learn to set up the claude python sdk, make api calls, handle responses, use system prompts, and stream output.
- shallow neural nets as smooth variational problems - a new formulation replaces discrete neural network training with a globally well-posed variational problem over parameter densities, enabling direct solution via a linear system.
- measuring lag relevance in time series forecasts - new model-agnostic method uses shapley values and ghost variables to measure lag importance in univariate time series forecasting.
- jersey mike's ipo shows ai hype has gone too far - a sandwich chain's ipo filing mentions ai 22 times, revealing how pervasive and hollow ai marketing has become even for non-tech businesses.
that's it for today. check back tomorrow for more updates on ai tools, research, and the occasional reality check.