Turning open-source funding into a token sounds like either a clever fix or a brand-new way to disappoint people at scale. And honestly, I’m not sure which outcome is more likely.
Jun Song, who’s well-known in the AI community, just launched something called the Supergemma token. From what’s been shared publicly, the goal is pretty straightforward: open-source builders are broke, the current ways of funding them don’t work well, and he thinks crypto mechanics can do a better job.
He also didn’t pretend everyone would cheer. He posted to his followers that people are confused and disappointed about him moving toward crypto. That part matters, because it tells you he knows he’s stepping into a trust problem, not just a product problem.
I actually agree with his core complaint. Traditional sponsorship models do tend to reward the same small set of visible winners. The people with a big name, a clean story, a popular repo, a loud fanbase. Meanwhile, a ton of useful work happens in the boring middle: maintenance, bug fixes, security patches, documentation, dependency upgrades. The stuff that keeps the whole machine running but doesn’t look sexy on a donation page.
If you’ve ever depended on a small library and then watched it break because the maintainer got a new job or burned out, you already understand the stakes. Your startup misses a deadline. Your team spends a week rewriting something you assumed would “just work.” A school project fails the night before it’s due. A volunteer maintainer gets blamed for not doing free labor fast enough. That’s the open-source economy we all quietly accept.
So yes, the system is lopsided. But tokenizing it is not automatically the solution. It might even make the incentives worse.
Here’s the tension: open-source needs steady, boring money. Tokens usually attract excited, impatient money. Open-source succeeds when people trust the work over time. Tokens often succeed when people believe the price will go up soon. Those are different moods. And moods shape behavior.
Imagine you’re a maintainer. You’re already juggling issues and pull requests after work. Now add a token community that wants “updates,” “roadmaps,” and “momentum.” Now your GitHub becomes half engineering and half investor relations. Even if nobody says it out loud, the pressure changes. You stop doing the unglamorous patch and start doing the flashy feature. You optimize for attention, because attention moves the token, and the token is supposed to fund you.
Or imagine you’re a user who just wants reliable software. You see a tool you depend on get wrapped in token talk. Maybe it’s fine. Or maybe you worry that the project will start making decisions for token optics instead of product quality. If trust slips, people leave. And open-source communities don’t collapse with a bang; they just slowly stop showing up.
The fairest case for what Jun is trying to do is this: if sponsorship mostly rewards a few stars, maybe a token can spread value across more contributors. Maybe it can reward small but important work. Maybe it can create a bigger pool of support than the same old “please sponsor me” button.
I get the appeal. I even respect the instinct to try something new instead of begging the same donors forever.
But I don’t buy the idea that crypto mechanics “revolutionize” funding without bringing the usual crypto problems along for the ride. Speculation isn’t a side effect in token systems; it’s often the main engine. That can attract money, sure. It can also attract the kind of behavior that makes builders miserable: hype cycles, tribalism, and endless arguments about fairness that turn into politics.
And then there’s the uncomfortable question of who wins first. In practice, early adopters tend to get the best deal in token projects. That doesn’t make it a scam, but it does mean you’re building an economy where timing matters a lot. If you’re a random maintainer who shows up late, do you get real support, or do you get crumbs while early holders benefit most? If you’re a newcomer trying to contribute, are you joining a community—or walking into a market?
Jun says people are disappointed, and I think that’s partly because open-source runs on a fragile social contract. People want to believe the work is about craft, curiosity, and public good. The moment money becomes tradable and public and tied to price, it changes the vibe. Not always for the worse, but it’s naive to act like it won’t.
What I’m genuinely unsure about is whether this token is designed to dampen speculation or to ride it. The summary says “crypto mechanics” can fix the funding problem, but it doesn’t say which mechanics, or how they prevent the usual mess. Without that, the idea sits in a dangerous middle zone: too financial to feel like pure community support, too idealistic to be treated like a normal investment.
If Supergemma works, it could pull more builders into sustainable work and make maintenance feel less like unpaid punishment. If it fails, it could make open-source feel like just another arena where attention and trading beat patience and care, and that would be a real loss for everyone who depends on these tools without even realizing it.
So what do we actually want open-source funding to optimize for: the most money possible, or the most stability possible?