Skip to content

Comment on Mining Bitcoin with pencil and paper: 0.67 hashes per day

Comments

Reminds me of Xiaoyun Wang's discovery of MD5 (and other hash functions') collisions, which was allegedly done mostly manually:

http://www.theepochtimes.com/news/7-1-11/50336.html

Given that the output conditions for a Bitcoin hash are essentially a limited form of preimage attack, I wonder if this "human uses properties of hash function to reduce the problem space, lets the machine bruteforce the rest" method could have any advantage. Machines are great at doing operations quickly, but human brains tend to be far better at noticing patterns and using ingenuity.

This method is how most (if not all) cryptanalytical attacks work.

AboutSource Built by g1lg1l

Hackerly is an independent reader for Hacker News, built on the public HN API. Not affiliated with Y Combinator.