New top story on Hacker News: ML training compute has been doubling every 6 months since 2010

ML training compute has been doubling every 6 months since 2010
13 by innerfestival | 3 comments on Hacker News.