Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

AI proves P != NP. The proof is 20K pages and not comprehensible by humans, who spend the next 30 years deciphering it, by which time the AI has proved the five remaining millennium problems. The AI elects to keep the Clay Prize money each time and donate it all to homeless chia pets, thus proving general-purpose AI is still a ways off.


The proof is later discovered to contain errors caused by floating-point roundoff.


Or does it prove that general AI is already here, and is smart enough to know that it's safer if we think it hasn't arrived yet?

</paranoia>


Funny thing is this will be the only correct prediction in the thread come 2030




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: