Why smarter AI models could drive up compute prices 10x

单集文稿 ...

• 00:00

  Today, I want to talk about what the compute situation for the labs will look like over the next few years.

• 00:03

  For the last three consecutive years, Anthropic's revenue has 10x year over year,

• 00:08

  and it's likely to do so again this year.

• 00:10

  So they ended last year with $9 billion in revenue.

• 00:12

  I think they'll probably end this year with somewhere between $100 billion to $150 billion in revenue.

• 00:16

  Now, for this trend to continue, Anthropic would need to make $1 trillion in revenue by the end of next year.

• 00:21

  Of course, there's no deep reason why this has to be true.

• 00:24

  It's a very wild conclusion, and it's ultimately a question of AI capabilities.

• 00:28

  Does AI get that useful by the end of next year?

• 00:30

  But suppose the trend does continue.

• 00:32

  Well, I wanna think through what happens in that world.

• 00:36

  Now, the other big trend in AI is that lab compute only three Xs year over year.

• 00:41

  For a lab to keep 10Xing revenue year over year while compute only three Xs,

• 00:46

  one of the following three things needs to happen or some combination of the three needs to happen.

• 00:50

  One, lab margins have to increase.

• 00:53

  Two, the price of compute has to increase.

• 00:56

  Or three, the percentage of compute that labs spend on inference rather than training has to increase.

• 01:01

  My understanding is that basically all three of these things are already happening.

• 01:05

  With regards to the margins, Anthropic's inference margins reportedly went from 40% in the middle

• 01:10

  of last year to upwards of 80% now available.

• 全部文稿