Research
Simon Willison
Visit site ↗📈
📈Business
Quoting Josh W. Comeau
Josh W. Comeau reports a significant decline in course sales, attributing it to AI's impact on the demand for learning new developer skills. Many people are hesitant to invest in courses due to concerns about AI replacing developer jobs, and those who still want to learn can use LLMs for personalized tutoring.
🔐
🔐Security
Fable's judgement
Simon Willison found a way to reduce costs with Fable by letting it use its own judgement to choose models for coding tasks, specifically delegating to lower-power subagents.
🚀
🚀Launches
June 2026 newsletter
GLM-5.2 was released, becoming the new best open weights model. This and other model releases were discussed in Simon Willison's June 2026 newsletter.
🚀
🚀Launches
llm-coding-agent 0.1a0
Simon Willison released llm-coding-agent 0.1a0, a coding agent built on LLM, which can write code in seconds and perform tasks like reading and editing files, and executing commands.
🔬
🔬Research
Using DSPy to evaluate and improve Datasette Agent's SQL system prompts
Simon Willison used DSPy to evaluate and improve Datasette Agent's SQL system prompts, finding potential improvements such as including column names in schema listings. This research aims to enhance the performance of Datasette Agent's read-only SQL question answerer.
🔬
🔬Research
Understand to participate
Geoffrey Litt emphasizes the importance of human understanding in collaborating with coding agents to avoid cognitive debt and participate effectively in the creative process.
You’re all caught up.