> A lot of ML systems run on Python. A lot of ML systems are first implemented in Python.
> That by being open source it doesn't deserve money? I think this is a flaw we probably need to fix.
Independent of how one feels about the current US administration, I do not think, as a non-American, that a particular government should foot the bill for it, but in reality I know that no company will do it in good will either.
I've been thinking a lot in terms of financing, but the current system of grants, where some agency tied with the executive body will approve or reject something, is fundamentally broken, as we can see.
In those cases of critical infrastructure, I think it's worth some kind of minimum 1:1 deductible of pre-tax programs where the foundations can apply, and then they could have their financing without being at the whims of some branch of the executive.
> I do not think, as a non-American, that a particular government should foot the bill for it
It is definitely a complicated problem but governments tend to but good funding agencies for work that uplifts the broader society and creates the foundation for new markets. That's the idea behind science funding anyways. New science might not create a trillion dollar business directly but it sure lays the funding for new multi billion dollar companies and companies to skyrocket from 500bn to 5T market caps...
But my point is that a project like this is global. I want the US putting money in. We're the richest and benefiting the most. But I also want other countries putting money in. They should have a vested interest too.
I think an interesting mechanism might be to use agencies like the NSA. We know their red teams but what about the blue? I'd love for the blue teams to get more funding and have a goal to find and patch exploits, rather than capitalize on them. Obviously should have a firewall between the teams. But this should be true for any country. It might just be some starting point as it could be a better argument for the people that don't already understand the extreme importance of these types of open source projects.
> I think it's worth some kind of minimum 1:1 deductible of pre-tax programs
Typically these projects run as nonprofit foundations. They're already getting tax benefits. Though I think we can recognize that this isn't enough and isn't remotely approaching the value.
It's definitely not an easy problem. Like what do you do? Tax big companies (idk, an extra 0.1%?), audit to determine dependencies, distribute those taxes accordingly? In theory this should be simple and could even be automated, but I'm sure in the cat and mouse game the complexity would increase incredibly fast.
But hey, it shouldn't just be America. Different countries can try different ideas
Independent of how one feels about the current US administration, I do not think, as a non-American, that a particular government should foot the bill for it, but in reality I know that no company will do it in good will either.
I've been thinking a lot in terms of financing, but the current system of grants, where some agency tied with the executive body will approve or reject something, is fundamentally broken, as we can see.
In those cases of critical infrastructure, I think it's worth some kind of minimum 1:1 deductible of pre-tax programs where the foundations can apply, and then they could have their financing without being at the whims of some branch of the executive.