Hacker News new | ask | show | jobs
by gwern 2656 days ago
> But the scale required to make a difference, and the cost to get to that scale boggles my mind. You can't just whip up one pack of mammoths and call it a success...

I assume the idea is that if you can create your big furry von Neumann machines correctly in the first place, you only need to whip up one pack of mammoths and it will indeed then be a success eventually.