scripod.com

Can AI Keep Growing Exponentially? Let’s Ask SemiAnalysis — With Dylan Patel and Jordan Nanos

Big Technology Podcast

Shownote

Dylan Patel is founder and CEO of SemiAnalysis, Jordan Nanos is a member of its technical staff. Patel and Nanos join Big Technology to discuss whether AI’s massive infrastructure buildout can justify the trillions of dollars being spent on compute. Tune ...

Highlights

Dylan Patel and Jordan Nanos examine whether AI infrastructure can keep expanding, how the buildout might be paid for, and what it means for cloud providers, frontier labs, workers, and the broader economy.
02:50
The real AI infrastructure numbers may be far higher
07:48
The AI buildout cannot accelerate forever
12:36
The payback can become clear
17:28
AI may accelerate the shift from labor to capital
20:32
AI could make knowledge work obsolete
24:37
Demand still outstrips supply
27:23
A $20 trillion Anthropic valuation could signal extreme AI power
30:49
Anthropic’s margins are already strong
39:44
NVIDIA is supporting a massive neocloud financing ecosystem
44:45
Many neoclouds are terrible at security
50:09
Oracle can be a great cloud with execution problems
57:29
Google wants every layer of the AI stack
1:02:42
Minimum viable models may define the next releases

Chapters

The scale and limits of the AI infrastructure boom
00:00
From industrial expansion to eventual payback
06:58
How AI creates the revenue to support its buildout
12:42
Knowledge work, neocloud risk, and Anthropic’s valuation
19:54
Why AI labs could become highly profitable
29:21
NVIDIA’s financing machine for neoclouds
36:11
Neocloud security and Oracle’s execution challenge
44:10
The uneven quality of AI clouds and Google’s strategy
52:47
Why frontier labs may keep their best models private
58:12

Transcript

Alex Kantrowitz: How big does AI have to get to justify the trillions in infrastructure spending? Let's talk about it with the leading authorities from Semi-Analysis right after this. This episode is brought to you by Genesys. What does it actually take to...