The shoulders of giants

Earlier this week, I wrote about a labor of love. Simon Cozens crafted a beautiful math font and shared the knowledge with all of us. Quite a gift of his time and talent.

We all benefit when people share broadly their ideas and techniques. Our careers and lives have been built on the shoulders of giants. Software lends itself to sharing better than almost any other technology; it is easy to share work and easy to create new derivative works.

Matthew Strom wrote a more extensive article on this – Copying (is the way design works) – which is worth a read. We learn from the works of others, and we, in turn, teach others and the next generation through our own work. This cycle of learning and teaching is what drives innovation and progress.

There are issues around citation and compensation for work. As it gets ever easier to copy work – as our tools encourage us through AI to leverage past work, sometimes unknowingly – we will have to create new ways to recognize precursors. But helping humans to better learn from and leverage past work seems like a good thing.

Mark Zuckerberg has articulated his plan to share Meta's AI work – Zuckerberg's Open Source AI letter. Some of this is a little self-serving – open source is a proven way to combat closed-source leaders. The license to use Llama also has substantial restrictions for large commercial users, so it isn't fully open source. It would be a tough decision to start building an ambitious commercial effort on the Meta tech, knowing that someday you would need to negotiate a license on unknown terms.

Mark also mentions global competitiveness issues in his letter:

The next question is how the US and democratic nations should handle the threat of states with massive resources like China. The United States’ advantage is decentralized and open innovation. Some people argue that we must close our models to prevent China from gaining access to them, but my view is that this will not work and will only disadvantage the US and its allies. Our adversaries are great at espionage, stealing models that fit on a thumb drive is relatively easy, and most tech companies are far from operating in a way that would make this more difficult. It seems most likely that a world of only closed models results in a small number of big companies plus our geopolitical adversaries having access to leading models, while startups, universities, and small businesses miss out on opportunities. Plus, constraining American innovation to closed development increases the chance that we don’t lead at all. Instead, I think our best strategy is to build a robust open ecosystem and have our leading companies work closely with our government and allies to ensure they can best take advantage of the latest advances and achieve a sustainable first-mover advantage over the long term.

Totally agree with all this. We have to assume that intelligent people in other countries are hard at work on this technology. Our openness and decentralization are huge assets; we need to let people innovate freely.

I built a little app using the prior Llama release to help me with my private data. I am anxious to update it and try again, and I am glad that Meta is making this available for all of us to bang against.