Towards a better Bitcoin: There are more useful "proof of work" algorithms out there...

It would be great to have a blockchain/crypto currency that had a more useful "proof of work" mechanism. Searching for hash solutions does not provide humanity with any practical side-benefits. It would be better to implement proof of work by solving hard optimization problems that can actually help humanity, such as the "protein folding" problem, and many others. That would have the side-effect of curing many diseases, feeding many people, predicting climate, and optimizing many industrial process.