Neurospicy explanation
Simon Willison is a developer who writes about software and AI and publishes tools including Datasette and LLM. His site is a place to follow explanations, examples and links back to the underlying work. If a new AI term has arrived before its explanation, pick one relevant article and work through what it actually demonstrates. Reading everything is not the assignment.
// Put it to use
A real-world example
You keep seeing the term “LLM” and want a concrete starting point rather than a list of things everyone apparently already knows.
- Read Simon’s About page to understand who is writing and the work behind the site.
- Open the linked LLM documentation and read what the tool is for.
- Choose one small usage example; identify its input, the model it uses and the output it produces.
- Write down what you now understand and the next question you still need answered.
What you getOne understood example and a useful next question, rather than twenty open tabs and a vague sense of being behind.
// The useful bits. The trade-offs.
Is it a good fit?
Curious developers who want detailed examples and source links.
Pros
- Explore technical ideas alongside project notes and working examples.
- Follow the original author and documentation instead of relying on a second-hand summary.
Cons & considerations
- Choose a starting topic; the site covers many technical subjects.
- Follow the source’s publication date and the version of any tool discussed.
A little more detail
Follow the source of an idea and give its author credit.
Simon Willison created Datasette and co-created Django. His own site brings together technical writing, project notes and links.
The related LLM usage guide is a concrete place to explore his work: it documents how a command-line tool accepts prompts and selects models. The original explanation stays with its author.
// Take the idea further
What else could you do?
Follow one question
Choose an article around something you are trying to understand and follow one useful source from it.
Learn through a small tool
Use the LLM documentation to trace one prompt from input to model to response.
References & credits4 sources +
Sources checked . Features and terms can change.