Simon Willison’s Weblog

Newsletters: 2023

1. 

   18 December 2023 · Substack

   The AI trust crisis

2. 

   10 December 2023 · Substack

   Datasette Enrichments: a new plugin framework for augmenting your data

3. 

   29 November 2023 · Substack

   llamafile is the new best way to run a LLM on your own computer

4. 

   22 November 2023 · Substack

   Deciphering clues in a news article to understand how it was reported

5. 

   15 November 2023 · Substack

   Exploring GPTs: ChatGPT in a trench coat?

6. 

   7 November 2023 · Substack

   ospeak: a CLI tool for speaking text in the terminal via OpenAI

7. 

   26 October 2023 · Substack

   Now add a walrus: Prompt engineering in DALL-E 3

8. 

   23 October 2023 · Substack

   Embeddings: What they are and why they matter

9. 

   14 October 2023 · Substack

   Multi-modal prompt injection image attacks against GPT-4V

10. 

    29 September 2023 · Substack

    Talking Large Language Models with Rooftop Ruby

11. 

    12 September 2023 · Substack

    Build an image search engine with llm-clip, chat with models with llm chat

12. 

    4 September 2023 · Substack

    LLM now provides tools for working with embeddings

13. 

    30 August 2023 · Substack

    Making Large Language Models work for you

14. 

    17 August 2023 · Substack

    Datasette Cloud, Datasette 1.0a3, llm-mlc and more

15. 

    3 August 2023 · Substack

    Catching up on the weird world of LLMs

16. 

    24 July 2023 · Substack

    sqlite-utils now supports plugins

17. 

    18 July 2023 · Substack

    Accessing Llama 2 from the command-line with the llm-replicate plugin

18. 

    12 July 2023 · Substack

    My LLM CLI tool now supports self-hosted language models via plugins

19. 

    19 June 2023 · Substack

    symbex: search Python code for functions and classes, then pipe them into a LLM

20. 

    8 June 2023 · Substack

    Understanding GPT tokenizers

21. 

    4 June 2023 · Substack

    It’s infuriatingly hard to understand how closed models train on their input

22. 

    27 May 2023 · Substack

    Lawyer cites fake cases invented by ChatGPT, judge is not amused

23. 

    18 May 2023 · Substack

    llm, ttok and strip-tags—CLI tools for working with ChatGPT and other LLMs

24. 

    12 May 2023 · Substack

    Delimiters won’t save you from prompt injection

25. 

    4 May 2023 · Substack

    Leaked Google document: “We Have No Moat, And Neither Does OpenAI”

26. 

    3 May 2023 · Substack

    Prompt injection explained, with video, slides, and a transcript

27. 

    1 May 2023 · Substack

    Enriching data with GPT3.5 and SQLite SQL functions

28. 

    25 April 2023 · Substack

    The Dual LLM pattern for building AI assistants that can resist prompt injection

29. 

    21 April 2023 · Substack

    What’s in the RedPajama-Data-1T LLM training set

30. 

    16 April 2023 · Substack

    Prompt injection: what’s the worst that can happen?

31. 

    12 April 2023 · Substack

    Running Python micro-benchmarks using the ChatGPT Code Interpreter alpha

32. 

    7 April 2023 · Substack

    We need to tell people ChatGPT will lie to them, not debate linguistics

33. 

    3 April 2023 · Substack

    Think of language models like ChatGPT as a “calculator for words”

34. 

    27 March 2023 · Substack

    AI-enhanced development makes me more ambitious with my projects

35. 

    24 March 2023 · Substack

    I built a ChatGPT plugin to answer questions about data hosted in Datasette

36. 

    19 March 2023 · Substack

    Could you train a ChatGPT-beating model for $85,000 and run it in a browser?

37. 

    13 March 2023 · Substack

    The Stable Diffusion moment for Large Language Models