Had to break for the night, but I know there's def something to this area you're alluding to. It's a valuable area to be playing in, since determining where research should be, that is really broken in my experience...!
I worked under someone named Puja Ohlhaver a few years ago, helping her run the first in what was supposed to be a series of funding round events. It was based on the topics discussed here:
The website and some code is here:
- https://www.pluralresearch.org/
- https://github.com/lexicongovernance/pluraltools-frontend
- https://github.com/PluralCC
The idea was to allow people who guess at the sites of future innovation (in semantic space) to gain more power to direct funds in each successive round. We ran the session at Eth.Berlin 2024, and gave away 50K EUR in grants i think.
The goal was kinda to reflow power from the funders and tenured researchers (who don't really follow the frontier) to the grad students (who have wider knowledge). A side-goal was also to promote cross-collaboration using Polis clusters (the tool I was managing).
Anyhow, it sounds like it might not be the exact same, but it's in your neighborhood -- I'll read the rest of the pitch tomorrow! You're in a valuable area though!
EDIT: I'm seeing some focus in other comments on the missing context of "the algorithm" and I think that's correct to point out, and an important omission. I've seen algorithms explained fully in public (see the talk above), and I'm not sure you should be operating from point of view of patenting it, because it's the whole magic, and it must be auditable, no? Besides, patenting an algorithm seems like unhelpful attempts at capture. I don't think I can really evaluate the pitch without knowing the algorithm, despite my thinking the opportunity is very ripe.