Neil's News

Halting Problem

30 June 2019

Raise the subject of Atlantis with an archeologist, and you'll get nothing but an exasperated sigh. Raise the subject of vaccines and autism with a doctor, and you'll get a similar response. Every field has got a small collection of questions which keep reappearing from the conspiratorial fringe, and no amount of work can ever make them go away.

Computer science is no exception. The "Halting problem" was proved to be unsolvable back in 1936 -- a decade before the first stored program computers were created. Despite this, there's an underground fringe that's devoted to attempting to prove that the halting problem can be proved.

Being a Google employee with some moderate visibility, I got one such manifesto in the mail a couple of months ago. For reference, here is the 57 page paper from the author's website.

My initial reaction, and that of my coworkers, was intrigue and an interest in finding the paper's flaw. However we were quickly disillusioned when the paper was, in the author's own words, written "from a broadly religious perspective". None of us could make any sense of the material, and we quickly gave up.

About a month later I got another letter wondering why none of the 50+ people the first letter was mailed to had replied.

< Previous | Next >

Legal yada yada: My views do not necessarily represent those of my employer or my goldfish.