- There is no effective monetization scheme to develop, maintain, improve, or redesign basic infrastructure that is critical and used by everybody
- There is lots of money thrown around for tiny little non-reusable services that depend on that infrastructure
Where and how money is spent on the core infrastructure is typically sporadic, subject to the whims or neglects of the big companies that have R&D budgets, and typically their best interest for the infrastructure is not quite that of the general public.
Example: the web, which grew from a simple consumer-is-in-control document sharing system to an incredibly complicated and misengineered producer-is-in-control application engine that abuses user data/privacy all throughout. Yes the infrastructure was developed, but very poorly and incrementally, and with bad intentions.
Yes the infrastructure is largely open-source / open-protocol, so that counts as a win? But I wish we could invent a society in which shared-infrastructure is valued and monetized proportional to its contribution to the society (which would outweigh the utility of the services on top of it), rather than vice-versa.
Open-source verifiable operating systems, security libraries: these are more important than your text-messaging app! But we don't know how to fund them.
Doesn't OpenSSL fit this definition perfectly?
...which is a rather unsettling property in the foundation of your pyramid.
I agree with your theory--I'm just sad to see that there isn't always a lot of that in practice.
Hell, they almost killed off SPARC support on Debian a month ago because of developer ADD.
I don't think of it as a "solved problem" though, so it'll be interesting to see where open source goes in the future. The public goods link I posted elsewhere here is a good read, although a lot of negativity regarding copyright has crept into it.