Simon Willison’s Weblog

Talking about Open Source LLMs on Oxide and Friends

I recorded an episode of the Oxide and Friends podcast on Monday, talking with Bryan Cantrill and Adam Leventhal about Open Source LLMs.

Financial sustainability for open source projects at GitHub Universe

I presented a ten minute segment at GitHub Universe on Wednesday, ambitiously titled Financial sustainability for open source projects.

LLM now provides tools for working with embeddings

LLM is my Python library and command-line tool for working with language models. I just released LLM 0.9 with a new set of features that extend LLM to provide tools for working with embeddings.

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

SemiAnalysis published something of a bombshell leaked document this morning: Google “We Have No Moat, And Neither Does OpenAI”.

Thoughts on AI safety in this era of increasingly powerful open source LLMs

This morning, VentureBeat published a story by Sharon Goldman: With a wave of new LLMs, open source AI is having a moment — and a red-hot debate. It covers the explosion in activity around openly available Large Language Models such as LLaMA—a trend I’ve been tracking in my own series LLMs on personal devices—and talks about their implications with respect to AI safety.

Working in public

I participated in a panel discussion this week for path to Citus Con, a series of Discord audio events that are happening in the run up to the Citus Con 2023 later this month.

Stanford Alpaca, and the acceleration of on-device large language model development

On Saturday 11th March I wrote about how Large language models are having their Stable Diffusion moment. Today is Monday. Let’s look at what’s happened in the past three days.

Support open source that you use by paying the maintainers to talk to your team

I think I’ve come up with a novel hack for the challenge of getting your company to financially support the open source projects that it uses: reach out to the maintainers and offer them generous speaking fees for remote talks to your engineering team.

Writing better release notes

Release notes are an important part of the open source process. I’ve been thinking about these a lot recently, and I’ve assembled some thoughts on how to do a better job with them.

How to build, test and publish an open source Python library

At PyGotham this year I presented a ten minute workshop on how to package up a new open source Python library and publish it to the Python Package Index. Here is the video and accompanying notes, which should make sense even without watching the talk.

Open source projects: consider running office hours

Back in December I decided to try something new for my Datasette open source project: Datasette Office Hours. The idea is simple: anyone can book a 25 minute conversation with me on a Friday to talk about the project. I’m interested in talking to people who are using Datasette, or who are considering using it, or who just want to have a chat.

Weeknotes: PG&E outages, and Open Source works!

My big focus this week was the PG&E outages project. I’m really pleased with how this turned out: the San Francisco Chronicle used data from it for their excellent PG&E outage interactive (mixing in data on wind conditions) and it earned a bunch of interest on Twitter and some discussion on Hacker News.

My JSK Fellowship: Building an open source ecosystem of tools for data journalism

I started a new chapter of my career last week: I began a year long fellowship with the John S. Knight Journalism Fellowships program at Stanford.

Datasette 0.28—and why master should always be releasable

It’s been quite a while since the last substantial release of Datasette. Datasette 0.27 came out all the way back in January.

sqlite-utils: a Python library and CLI tool for building SQLite databases

sqlite-utils is a combination Python library and command-line tool I’ve been building over the past six months which aims to make creating new SQLite databases as quick and easy as possible.

What are some good open source projects that VMware is directly part of?

To my knowledge they fund almost all of the development work on RabbitMQ, Redis and the Spring Java framework.

What are the most commonly used or most interesting open-source packages and software?

I’d say the open source browser engines, Gecko (Firefox) and WebKit (Safari, Chrome, iOS, Android) are probably some of the most important and widely used pieces of open source code these days.

To become a better developer ? To read more OR to create/contribute to open source projects?

Contribute to an existing project, rather than starting one yourself. There are a bunch of benefits:

Is there a free/open-source software source code search engine?

If you want to search through actual code in open source projects, GitHub search is fantastic https://github.com/search—e.g. here’s a search for all Ruby code that mentions oauth https://github.com/search?q=oaut...

How can a new developer get involved in open-source projects?

I was going to say the same thing. Find a useful project in GitHub (preferably one that clearly has an active maintainer), fork it, fix a bug (look at the project’s issue tracker) then make a pull request.

What is the typical time overhead in using an open source package vs. an equivalent commercial package?

It totally depends on the software in question. If you are moving to popular, well maintained open source packages (things like nginx, solr, MySQL) you can often expect a large improvement in developer productivity due to the increased amount of tutorials, forums, mailing lists, irc channels and stackoverflow/quora posts you’ll have access to.

What is a good business model for open source projects?

Consulting.

What is the story of Advogato?

There’s a Google Tech Talk about Advogato: http://video.google.com/videopla...

Why I like Redis

I’ve been getting a lot of useful work done with Redis recently.

Announcing dmigrations

The team at Global Radio (formerly GCap Media) is the largest group of Django developers I’ve personally worked with, consisting of 14 developers split into two scrum teams, all contributing to the same overall codebase.

What are some good software development/open source blogs?

http://www.thedailywtf.com/ is an endlessly entertaining guide to how not to develop software.

Taking charge of your own destiny

Scoble has posted 12 reasons that Web 2.0 entrepreneurs are steering clear of the Microsoft platform. It’s an interesting list (the comments are full of treats too) but for me it misses the key reason that open source development tools are so compelling: they put you in charge of your own destiny.

Enter the hedgehog

The Ubuntu community have released Hoary Hedgehog, otherwise known as Ubuntu 5.04. If you haven’t tried Ubuntu yet, it’s an excellent Linux distribution based on Debian with a strong focus on desktop usability. Unlike most Linux distros, Ubuntu comes with just one desktop manager (Gnome) and one obvious default application for each of the essentials: Firefox for browsing, OpenOffice for office work, Evolution for mail.

The good and the ugly

PHP.net has a new feature on their search page—a really nice implementation of an auto complete text widget in Javascript. Even better, the search page is valid XHTML 1.0 Strict and uses CSS for the layout. Let’s hope this is an indication of things to the come for the rest of the site, which still mostly consists of tag soup.

Clearout

• Tristan Louis’ RSS to Necho convertor puts paid to the idea that the success of one format will be detrimental to the usefulness of the other.
• O’Reilly’s RegExp Power series (part one and part two) demonstrate some powerful tricks for use with Perl compatible regular expressions.
• Norman Walsh explains Content Negotiation and some of the pitfalls with modern browser implementations.
• So that’s what happened to Digitiser. See also a Digitiser Tribute and a Mr Biffo interview from 2001 for background information. I cuss you bad.
• George Orwell: Politics and the English Language
• Clay Shirky: A Group Is Its Own Worst Enemy. The title is misguiding; this is an essay about how online groups behave and how to look after them.
• A Java HttpClient Class.
• Some good stuff on Boxes and Arrows: Ten Quotable Moments: Challenges and Responses for UI Designers and Views and Forms: Principles of Task Flow for Web Applications (Part 1).
• Inside our notions of “document” and Inside our documents II—the Runoff model.
• 5 days worth of XSLT observations from Simon St. Laurent: One, Two, Three, Four, Five.
• Windows programming with open source tools: Minimalist GNU For Windows and Win32 Programming with GNU C and C++.