Re:Microsoft's Quake 2 AI experiment sparks negative reactions
by kirtakat on Sunday April 06, 2025 @01:51PM
(#43603351)
Currently this is a fever dream of non-Euclidean space - but the fact that it can reliably do this at all is extremely impressive.
[ Parent
| on HN ]
Re:Microsoft's Quake 2 AI experiment sparks negative reactions
by timewizard on Sunday April 06, 2025 @02:02PM
(#43603466)
> but the fact that it can reliably do this at all is extremely impressive
The better question is could you achieve the same thing using much less computing power and simpler methods? Would a Markov chain be able to do the same thing? If so, I'd have to with the articles sentiment, this burns a bunch of electricity to do something with extreme inefficiency and waste.
Re:Microsoft's Quake 2 AI experiment sparks negative reactions
by jsheard on Sunday April 06, 2025 @02:22PM
(#43603608)
> The better question is could you achieve the same thing using much less computing power and simpler methods?
If only there were a way to play Quake 2 on a 90mhz Pentium with 16MB of RAM...