Focus on things that can't be funded in traditional academia, yet which are public goods and so cannot be funded by startups. Project that have one or more of

Focus on things that can't be funded in traditional academia, yet which are public goods and so cannot be funded by startups. Project that have one or more of the following characteristics: 1) requires a large amount of implementation work and real-world trials before you know if it's a good idea 2) a long project, but if it fails, there may be no publishable intermediate results (in academia, we stay away from long projects unless they are publishable no matter what the result turns out to be; b/c at the end of grad school, or your postdoc, or your tenure clock, you need a publication in order to remain in academia) 3) reform of academic funding (it's not advisable to pitch an NSF government official to fund a project that only makes sense if NSF government officials are unnecessary) 4) success criterion is a matter of taste, rather than objective (or, maybe the true success criterion is objective but too expensive to measure, so taste must be substituted as a proxy success criterion) 5) creation of useful artifacts, rather than new knowledge 6) requires ongoing maintenance in order to remain useful to society 7) more a matter of putting together or implementing existing ideas in a wholistic way than discovery of novel ideas Examples: * alternative systems to allocate research funding (satisfies #1,3,4) * creation of new programming languages (1, 2, 4, 5, 6, 7) * open-source software development (5, 6) * creation of new software-supported governance procedures (1, 4, 7) * creation of new legal structures (for remote working open co-ops, etc) (1, 4, 5) * direct attacks on the Millennium problems in math (1) * upgrades to the academic peer review process (1, 4, 6) * alternative systems for hiring professors (1, 3, 4)