←back to thread

129 points celias | 1 comments | | HN request time: 0.21s | source
Show context
rfv6723 ◴[] No.44393382[source]
Apple AI team keeps going against the bitter lesson and focusing on small on-device models.

Let's see how this would turn out in longterm.

replies(5): >>44393454 #>>44393509 #>>44393622 #>>44394586 #>>44394727 #
sipjca ◴[] No.44393509[source]
somewhat hard to say how the cards fall when the cost of 'intelligence' is coming down 1000x year over year while at the same time compute continues to scale. the bet should be made on both sides probably
replies(1): >>44393733 #
furyofantares ◴[] No.44393733[source]
10x year over year, not 1000x, right? The 1000x is from this 10x observation having held for 3 years.
replies(1): >>44416258 #
1. sipjca ◴[] No.44416258[source]
I believe the 1000x number I pulled is from SemiAnalysis or similar, using MMLU as the baseline benchmark and the cost per token from a year ago to today at the same score. Model improvements, hardware improvements and software improvements all make a massive difference when combined to make much greater than 10x gains in terms to intelligence/$