My friend Konstantin Vinogradov is launching something new — an Open Source Endowment (OSE), a US-based nonprofit that will fund open-source work that silently powers the world. I've joined in. Maybe you should, too 🤷🏻
Over the last 12 months, I have reached out to several tech companies worth billions/trillions — ones that use my open-source code in production — to explore more sustainable options for working on open-source projects. On one hand, some top AI labs had offered me 8-digit contracts to work on their internal infrastructure. On the other hand, I was asking for just $20,000–$50,000 from each of 5-10 organizations to help maintain the code they already use in production. I thought it would be a no-brainer. But everyone politely passed.
Moreover, the cognitive load of those conversations was very distracting from the actual R&D. Individual efforts didn't scale. And built-in features, like "GitHub Sponsors", don't work either. A more organized approach is needed! An endowment?
Conveniently, the contributions don't have to be huge. Countless developers I know would pay $200/month or $2,000/year for each of their LLM subscriptions for coding. The output of those LLMs is very emotionally supportive for me when debugging complex issues, but from a practical perspective, 90% of LLM-generated code is an ugly & broken mess, 5% is a pure replica of somebody else's open-source work, and 5% involves novel interpolation between (again!) existing public examples.
With such a weak baseline, many techies should be open to donating $1,000 per year. With a million individuals, even excluding corporate donations, that's a total of $1 billion — enough to pay 5,000 developers $200,000 per year. Almost double the headcount of OpenAI. A sizable budget that can make a huge difference. Dreamy, but not impossible!
My preference lies with the forgotten art of building compact low-level infrastructure projects, such as the Tiny C Compiler, OpenBLAS, and JeMalloc. The current draft of the distribution plan doesn't fully align with such preferences, but it's a good start, and I'd love to see this succeed!
There are already some other amazing people involved, including Chad Whitacre (Sentry), Maxim Konovalov (NGINX), and Jonathan Starr (NumFocus). OSE is also advised by Amy Parker, CFRE (OpenSSL Foundation, former Wikimedia) and Vlad-Stefan Harbuz (Open Source Pledge). What other names do you think should be on that list?
https://endowment.dev
https://lnkd.in/gQ8NHjTA