The Innovation Game: The Market-Driven Approach to Solving Big Science Problems

You're now plugged in to the Delphi podcast, my god. Welcome to Delphi podcast. I'm John from Delphi Ventures. And today with me, I have doctor John Fletcher, the CEO and cofounder of the Innovation Game. The Innovation Game, TIG, is a decentralized science project, which frankly is one of the most uniquely innovative projects I've seen in crypto in recent years. In one sentence, I can describe TIG as a coordination protocol designed to accelerate algorithmic innovation. The TIG network, through useful people work, discovers the most efficient algorithm for a given scientific problem and conti

very much. Good to be here. John,

given how big of a project TIG is, there's a lot to unpack in this episode. Why don't we start with a high level introductory overview of what TIG is in your own words? Briefly describe us what what TIG is and what is its purpose.

Yeah. Sure. So, TIG is, stands for the innovation game, and, its purpose is to accelerate, open algorithmic innovation. So these algorithms can be anything from, you know, that, an algorithm which turns up measurements of a magnetic field into an image on an MRI machine to an algorithm which sort of minimizes a cost function in a in an AI training run. The sort of algorithms are really everywhere. And the the sort of key insight that we had is that the the sort of open development of these algorithms is not really sort of as successful as, say, the most successful open source projects. They re

Perfect. Thank you very much, for the intro. There's definitely, like, components that I wanna dive deeper here. But I'm actually first curious on, like, it's a it's a truly kind of unique project, and I'm I'm curious to hear the origin story of TIG. How did you come up with this idea? And since when you've been working on it? Yeah.

Great question. I I I I should go and check this back in my my archives, but I think I had the first I had the idea in in 2012, so quite a long time ago. And sort of back then, I was, I was a PhD student, in fact, in the in the math department in Cambridge. I was lucky enough to have an office which was, just opposite Stephen Hawking's office. So I hear kind of like not his door, but his but the window to his office so I could sort of see in. And, one of the things about professor Hawking is that you wouldn't really, know that he worked very long hours. He kind of got there quite early, stayed

Very interesting. I wanna double click on the on the problem that you were you were trying to solve. That's it's kinda interesting to hear. Could could you maybe briefly explain what it was? Yeah.

Right. So, you have, sort of many, surfaces and, kind of materials in nature have a sort of fractal property at least to some extent. So sometimes when you zoom in very close, they sort of stop becoming fractals, or maybe when you sort of zoom out enough, but over a kind of a large range of scales, they are sort of fractals. And that one kind of intuitive way to think about this is, yeah, sometimes if you see a picture of some rocks, like, you can't really tell whether you're flying over it in a plane and it's mountain range or whether you you have a microscope and you're looking at it and the

And then and you were working on this, basically, solution, which, and you realize, like, the the nature of the solution is asymmetric. You have to spend a lot of Yeah. The