AI researchers always disappoint me because I haven't ever seen a discussion of outright tampering. The danger of AI isn't free will, but some suicidal maniac c

AI researchers always disappoint me because I haven't ever seen a discussion of outright tampering. The danger of AI isn't free will, but some suicidal maniac capable of embedding viruses in the AI. We've got enough of black-hat hackers around—my employer got hit by cryptolocker last year—who could come up with a way to profit by hacking AIs.