I was excited about the possibility of combining BOINC projects with cryptocurrency, which could result in magnitudes more computational power being contributed to important mathematical and scientific problems through the built-in economic incentivisation mechanisms of cryptocurrencies.
I have been developing cryptocurrencies since 2013, and I think it may in fact be possible to do this. The first project of this kind was called Primecoin ( http://primecoin.io/ ) which economically incentivizes finding long Cunningham Chains. For several years, this project found some of the largest known cunningham chains.
It would be exciting to see if cryptocurrency incentivisation can also be applied to BOINC project. For it to work, there are 2 requirements:
1) The mathematical problem must be divisible into smaller subproblems that can allow contributors of computational power to "prove" that they in fact contributed computational power to the BOINC project.
2) The solution to the smaller subproblems must be "easy to verify" by anyone. This means that it shouldn't take longer than a minute for a typical computer to verify that someone in fact did the "computational work" needed to get the solution to the smaller subproblem. (basically, the same concept as "proof of work" found in Bitcoin)
Note: There is already a project called "Gridcoin" that contributes computational power directly to BOINC. However, Gridcoin requires a central party to verify that the computational work was actually done, so it is not a truly distributed cryptocurrency. This is why the 2 above requirements are important.
Cryptocurrency can be combined with any mathematical or scientific problem, not just strictly pre-existing BOINC projects. If anybody has ideas, I'd love to hear your thoughts :)