> Humans can do mistakes. Machines can't do mistakes (unless humans who programmed them did mistakes) That's an incredibly limited view of the issue. Machines

> Humans can do mistakes. Machines can't do mistakes (unless humans who programmed them did mistakes) That's an incredibly limited view of the issue. Machines can make mistakes regardless of human intervention (albeit probably indirectly, because humans have an unprecedented level of influence on reality). First, there are known software glitches caused by unexpected bit flips. Second, software systems can grow to a level of complexity where unless you invest orders of magnitude more time in theorem provers for it you cannot guarantee that 'machines can't do mistakes'. Electronic voting is an interesting problem and I agree with you, it's probably the future. But it's not a near future or not as near as you'd think. Electronic voting is vulnerable to attacks which cannot be detected when they happen. Armies of diverse human observers for paper ballots are much more effective for detecting fraud.