I've been leaning towards the conclusion that better innovations in education (this includes both technology and cultural/institutional) would be a huge, multiplicative high level action that would benefit most other initiatives people favor. I strongly doubt we've come close to exhausting the low-hanging fruit in education technique (and in the flexibility to apply techniques as needed to different types of people)
Programs like KIPP applied one major overhaul to the existing system... and then stopped. I'd like to donate to an organization that systematically tries out radically different systems and technological innovations and then attempts to replicate the ones that work.
Improved ways to educate people would indeed be a very worthy goal; sadly, there seems to be little consensus (let alone controlled experiments) as to what makes a good education to whom. Is it the quality of teachers, the size of classes, the availability of technology (or the lack thereof), the homogeneity (or heterogeneity) of the subjects-to-be-taught, rigidity of the school system (or freedom the follow one's own path), etc. etc. that help teaching/learning the most? The research points in opposite ways, and what actually happens in practice seems, to...
I have just received a survey questionnaire regarding future directions in EU (European Union) research funding, and thought it would be interesting to see how LessWrong would answer the main question:
Imagine that EU funding is available for one ambitious, visionary project extending beyond 2020.