clip |
https://simonwillison.net/tags/clip/ |
[CLIP](https://github.com/openai/CLIP) is an embedding model by OpenAI that supports comparisons between images and text. |
8 |
2 |
6 |
0 |
annotated-release-notes |
https://simonwillison.net/tags/annotated-release-notes/ |
I sometimes publish extended annotated versions of the release notes for my projects. |
24 |
22 |
2 |
0 |
evals |
https://simonwillison.net/tags/evals/ |
Evaluations and benchmarks of AI systems, for example to compare different models and prompting strategies. |
11 |
1 |
8 |
2 |
marimo |
https://simonwillison.net/tags/marimo/ |
[Marimo](https://marimo.io/) is an open-source reactive notebook for Python. |
4 |
0 |
4 |
0 |
weeknotes |
https://simonwillison.net/tags/weeknotes/ |
I publish weeknotes once every two or three weeks documenting what I've been working on recently. |
191 |
191 |
0 |
0 |
datasette-cloud |
https://simonwillison.net/tags/datasette-cloud/ |
The SaaS hosted version of Datasette, at [datasette.cloud](https://www.datasette.cloud/) |
45 |
33 |
12 |
0 |
s3-credentials |
https://simonwillison.net/tags/s3-credentials/ |
[s3-credentials](https://s3-credentials.readthedocs.io/) is a CLI tool for creating and managing credentials for S3 buckets. |
9 |
7 |
2 |
0 |
llm-tool-use |
https://simonwillison.net/tags/llm-tool-use/ |
Tool use is when an LLM is instructed to occasionally request that an external tool be run on its behalf, with the result passed back to the model for further processing. Sometimes also known as function calling, and one of several ideas that might be referred to as "agents". |
12 |
2 |
10 |
0 |
ham-radio |
https://simonwillison.net/tags/ham-radio/ |
The [amateur radio](https://en.wikipedia.org/wiki/Amateur_radio) hobby. |
3 |
0 |
3 |
0 |
phi |
https://simonwillison.net/tags/phi/ |
Microsoft's [Phi family](https://azure.microsoft.com/en-us/products/phi) of small Large Language Models. |
7 |
1 |
6 |
0 |
llm |
https://simonwillison.net/tags/llm/ |
[LLM](https://llm.datasette.io/) is my command-line tool for running prompts against Large Language Models. |
120 |
48 |
70 |
2 |
recall |
https://simonwillison.net/tags/recall/ |
Microsoft Recall is a feature planned for Windows 11 that allows users to search through automated screenshots taken of their previous activity on their PC. |
4 |
0 |
3 |
1 |
rag |
https://simonwillison.net/tags/rag/ |
RAG stands for **Retrieval Augmented Generation**. It's a trick where you find additional context relevant to the user's request using other means (such as full-text or vector search) and populate that context as part of the prompt to a Large Language Model. |
27 |
8 |
15 |
4 |
projects |
https://simonwillison.net/tags/projects/ |
Posts about projects I have worked on. |
449 |
248 |
201 |
0 |
datasette |
https://simonwillison.net/tags/datasette/ |
[Datasette](https://datasette.io/) is an open source tool for exploring and publishing data. |
422 |
235 |
185 |
2 |
claude |
https://simonwillison.net/tags/claude/ |
[Claude](https://www.anthropic.com/claude) is Anthropic's family of Large Language Models. |
108 |
23 |
70 |
15 |
anthropic |
https://simonwillison.net/tags/anthropic/ |
[Anthropic](https://www.anthropic.com/) are the AI research company behind [Claude](https://claude.ai/). |
103 |
22 |
63 |
18 |
llms |
https://simonwillison.net/tags/llms/ |
Large Language Models (LLMs) are the class of technology behind generative text AI systems like OpenAI's [ChatGPT](https://simonwillison.net/tags/chatgpt/), Google's [Gemini](https://simonwillison.net/tags/gemini/) and Anthropic's [Claude](https://simonwillison.net/tags/claude/). |
834 |
133 |
509 |
192 |
openai |
https://simonwillison.net/tags/openai/ |
[OpenAI](https://openai.com/) build [ChatGPT](https://chatgpt.com/) and the GPT series of Large Language Models. |
232 |
59 |
135 |
38 |
mistral |
https://simonwillison.net/tags/mistral/ |
[Mistral AI](https://mistral.ai/) release both openly licensed and API-hosted Language Models. |
20 |
4 |
15 |
1 |
boring-technology |
https://simonwillison.net/tags/boring-technology/ |
[boringtechnology.club](https://boringtechnology.club/) |
7 |
3 |
4 |
0 |
datasette-lite |
https://simonwillison.net/tags/datasette-lite/ |
Datasette running entirely in the browser using Pyodide and WebAssembly, at [lite.datasette.io](https://lite.datasette.io/) |
16 |
12 |
4 |
0 |
sqlite-utils |
https://simonwillison.net/tags/sqlite-utils/ |
[sqlite-utils](https://sqlite-utils.datasette.io/) is a Python library and command-line utility for creating and manipulating SQLite databases. |
86 |
60 |
26 |
0 |
shot-scraper |
https://simonwillison.net/tags/shot-scraper/ |
[shot-scraper](https://shot-scraper.datasette.io/) is a command-line utility for taking screenshots of websites and scraping content from them using JavaScript. |
25 |
13 |
12 |
0 |
training-data |
https://simonwillison.net/tags/training-data/ |
Data used to train LLMs and other machine learning models. |
45 |
8 |
25 |
12 |
sqlite |
https://simonwillison.net/tags/sqlite/ |
[SQLite](https://www.sqlite.org/) is the world's [most widely deployed database engine](https://www.sqlite.org/mostdeployed.html). |
285 |
75 |
195 |
15 |
embeddings |
https://simonwillison.net/tags/embeddings/ |
See [Embeddings: What they are and why they matter](https://simonwillison.net/2023/Oct/23/embeddings/). |
45 |
7 |
38 |
0 |
nomic |
https://simonwillison.net/tags/nomic/ |
[Nomic AI](https://www.nomic.ai) develop various interesting AI projects, including [GPT4All](https://gpt4all.io/) and powerful [embedding models](https://gpt4all.io/). |
5 |
0 |
5 |
0 |
python |
https://simonwillison.net/tags/python/ |
The [Python](https://www.python.org/) programming language. |
1079 |
191 |
860 |
28 |
o1 |
https://simonwillison.net/tags/o1/ |
OpenAI’s [o1 family](https://openai.com/o1/) of models. |
15 |
5 |
4 |
6 |
psf |
https://simonwillison.net/tags/psf/ |
The [Python Software Foundation](https://www.python.org/psf-landing/) is the non-profit organization devoted to advancing open source technology related to the Python programming language. |
12 |
4 |
8 |
0 |
dsf |
https://simonwillison.net/tags/dsf/ |
The [Django Software Foundation](https://www.djangoproject.com/foundation/). |
10 |
1 |
8 |
1 |
ladybird |
https://simonwillison.net/tags/ladybird/ |
[Ladybird](https://ladybird.org/) is "a truly independent web browser". |
5 |
0 |
4 |
1 |
beeware |
https://simonwillison.net/tags/beeware/ |
[BeeWare](https://beeware.org/) "allows you to write your app in Python and release it on multiple platforms". |
3 |
0 |
3 |
0 |
vs-code |
https://simonwillison.net/tags/vs-code/ |
Microsoft's partially open source text editor [Visual Studio Code](https://code.visualstudio.com/). |
4 |
0 |
4 |
0 |
markdown-exfiltration |
https://simonwillison.net/tags/markdown-exfiltration/ |
**Markdown Exfiltration** is a [prompt injection](https://simonwillison.net/tags/promptinjection/) attack against chatbots that have access to private information and can render Markdown images. An attacker can trick the bot into rendering an image that leaks data encoded in the URL to an external server. |
15 |
3 |
11 |
1 |
val-town |
https://simonwillison.net/tags/val-town/ |
[Val Town](https://www.val.town/) provides a web interface for building, running and deploying server-side JavaScript apps. |
6 |
1 |
5 |
0 |
prompt-injection |
https://simonwillison.net/tags/prompt-injection/ |
**Prompt Injection** is a security attack against applications built on top of Large Language Models, [introduced here](https://simonwillison.net/2022/Sep/12/prompt-injection/) and further described in [this series of posts](https://simonwillison.net/series/prompt-injection/). |
83 |
23 |
49 |
11 |
overture |
https://simonwillison.net/tags/overture/ |
[Overture](https://overturemaps.org/) provides interoperable open map data. |
6 |
1 |
5 |
0 |
edge-llms |
https://simonwillison.net/tags/edge-llms/ |
LLMs that can run on consumer hardware like laptops or mobile phones. |
67 |
18 |
47 |
2 |
llm-pricing |
https://simonwillison.net/tags/llm-pricing/ |
Posts about the pricing of various LLMs. See also my [pricing calculator](https://tools.simonwillison.net/llm-prices). |
15 |
4 |
11 |
0 |
smollm |
https://simonwillison.net/tags/smollm/ |
[SmolLM](https://huggingface.co/blog/smollm) is a series of small Large Language Models by Hugging Face. |
4 |
0 |
4 |
0 |
structured-extraction |
https://simonwillison.net/tags/structured-extraction/ |
Using LLMs to extract structured data from unstructured text and images. |
6 |
1 |
5 |
0 |
pelican-riding-a-bicycle |
https://simonwillison.net/tags/pelican-riding-a-bicycle/ |
My [benchmark for LLMs](https://simonwillison.net/2024/Oct/25/pelicans-on-a-bicycle/): "Generate an SVG of a pelican riding a bicycle" |
11 |
3 |
8 |
0 |
claude-artifacts |
https://simonwillison.net/tags/claude-artifacts/ |
A feature of [Claude](https://claude.ai/) where it can create shareable, executable HTML and JavaScript pages. See [Everything I built with Claude Artifacts this week](https://simonwillison.net/2024/Oct/21/claude-artifacts/) for several detailed examples. |
24 |
8 |
15 |
1 |
cors |
https://simonwillison.net/tags/cors/ |
CORS is [Cross-origin resource sharing](https://developer.mozilla.org/en-US/docs/Web/HTTP/CORS), a mechanism by which websites can make APIs available to JavaScript running on other domains around the web. |
20 |
9 |
10 |
1 |
pydantic |
https://simonwillison.net/tags/pydantic/ |
[Pydantic](https://docs.pydantic.dev/) is a data validation library for Python based around Python's type annotation feature. |
9 |
1 |
7 |
1 |
ollama |
https://simonwillison.net/tags/ollama/ |
[Ollama](https://ollama.com/) is a tool for downloading and running Large Language Models. |
11 |
3 |
8 |
0 |
inference-scaling |
https://simonwillison.net/tags/inference-scaling/ |
Improving performance of LLMs through inference - spending more tokens "thinking" about a problem, as seen in OpenAI's o1 and Qwen's QwQ. |
10 |
4 |
4 |
2 |
code-interpreter |
https://simonwillison.net/tags/code-interpreter/ |
ChatGPT Code Interpreter is a mode of [ChatGPT](https://chatgpt.com/) that allows it to write and then execute Python code in a sandboxed environment. OpenAI sometimes call this feature [data analysis](https://help.openai.com/en/articles/8437071-data-analysis-with-chatgpt). |
20 |
12 |
8 |
0 |
chrome |
https://simonwillison.net/tags/chrome/ |
Google's [Chrome](https://www.google.com/chrome/) browser. |
30 |
3 |
20 |
7 |
slop |
https://simonwillison.net/tags/slop/ |
**Slop** describes AI-generated content that is both *unrequested* and *unreviewed*. See [Slop is the new name for unwanted AI-generated content](https://simonwillison.net/2024/May/8/slop/). |
14 |
2 |
10 |
2 |
llamafile |
https://simonwillison.net/tags/llamafile/ |
[llamafile](https://github.com/Mozilla-Ocho/llamafile) provides a mechanism to distribute and run LLMs as a single file. |
6 |
4 |
2 |
0 |
django |
https://simonwillison.net/tags/django/ |
The [Django](https://www.djangoproject.com/) web framework. |
566 |
100 |
453 |
13 |
go |
https://simonwillison.net/tags/go/ |
The [Go](https://go.dev/) programming language. |
36 |
4 |
29 |
3 |
ai-assisted-programming |
https://simonwillison.net/tags/ai-assisted-programming/ |
Using AI tools such as Large Language Models to help write code. |
98 |
31 |
58 |
9 |
command-r |
https://simonwillison.net/tags/command-r/ |
The [Command R](https://cohere.com/command) family of language models from Cohere. |
2 |
0 |
2 |
0 |
github-codespaces |
https://simonwillison.net/tags/github-codespaces/ |
[GitHub Codespaces](https://github.com/features/codespaces) provides cloud development environments accessible through a browser. |
10 |
3 |
7 |
0 |
annotated-talks |
https://simonwillison.net/tags/annotated-talks/ |
Annotated versions of talks I have given, with extensive notes and additional links. Here's [how I make these](https://simonwillison.net/2023/Aug/6/annotated-presentations/). |
23 |
23 |
0 |
0 |
datasette-desktop |
https://simonwillison.net/tags/datasette-desktop/ |
Datasette as an Electron desktop application, at [datasette.io/desktop](https://datasette.io/desktop) |
6 |
5 |
1 |
0 |
vision-llms |
https://simonwillison.net/tags/vision-llms/ |
LLMs that can also be used to interpret images and video, such as GPT-4o, Claude 3 and Gemini Pro. |
34 |
8 |
25 |
1 |
prompt-engineering |
https://simonwillison.net/tags/prompt-engineering/ |
The subtle art and craft of effectively prompting and building software on top of LLMs. |
112 |
30 |
61 |
21 |
github-copilot |
https://simonwillison.net/tags/github-copilot/ |
[GitHub Copilot](https://github.com/features/copilot) is GitHub's extension for VS Code that uses LLMs to provide context-driven code completion suggestions. |
13 |
4 |
8 |
1 |
github-actions |
https://simonwillison.net/tags/github-actions/ |
GitHub's [Actions](https://github.com/features/actions) tool for repository automation. |
47 |
25 |
22 |
0 |
git-history |
https://simonwillison.net/tags/git-history/ |
[git-history](https://github.com/simonw/git-history) is a tool for analyzing data scraped to a GitHub repository using SQLite. |
6 |
5 |
1 |
0 |
djp |
https://simonwillison.net/tags/djp/ |
[DJP](https://djp.readthedocs.io/) is a plugin system for Django. |
5 |
2 |
3 |
0 |
git-scraping |
https://simonwillison.net/tags/git-scraping/ |
[Git scraping](https://simonwillison.net/2020/Oct/9/git-scraping/) is a technique where data is scraped from an external source into a Git repository in order to record changes to that data over time. |
29 |
22 |
7 |
0 |
django-sql-dashboard |
https://simonwillison.net/tags/django-sql-dashboard/ |
[Django SQL Dashboard](https://django-sql-dashboard.datasette.io) adds a read-only PostgreSQL SQL dashboard interface to a Django project. |
13 |
8 |
5 |
0 |
mlx |
https://simonwillison.net/tags/mlx/ |
The [MLX framework](https://github.com/ml-explore/mlx) for running machine learning models on Apple Silicon. |
6 |
2 |
4 |
0 |
notebooklm |
https://simonwillison.net/tags/notebooklm/ |
Google's [NotebookLM](https://notebooklm.google/) "personalized research assistant". |
5 |
1 |
2 |
2 |
observable-plot |
https://simonwillison.net/tags/observable-plot/ |
The <a href="https://observablehq.com/plot/">Observable Plot</a> JavaScript visualization library. |
9 |
1 |
8 |
0 |
gpus |
https://simonwillison.net/tags/gpus/ |
GPUs aren't just for graphics anymore. |
12 |
0 |
9 |
3 |
ai-agents |
https://simonwillison.net/tags/ai-agents/ |
I don't think the term "agents" has a useful, widely shared definition. |
14 |
3 |
7 |
4 |