Hmm I think Leonid Levin did something similar: https://www.cs.bu.edu/fac/lnd/expo/holo.htm https://www.cs.bu.edu/fac/lnd/expo/holo.htm Not sure yet what this

Hmm I think Leonid Levin did something similar: https://www.cs.bu.edu/fac/lnd/expo/holo.htm https://www.cs.bu.edu/fac/lnd/expo/holo.htm Not sure yet what this one is doing that that one doesn't. (Edited to add) Having read further, it seems that this code is close to optimal in rate and distance too. The link above is actually for probabilistically checkable proofs, which are similar, and the article thinks this is likely to lead to advances there too