Eisman: 70% of hyperscaler AI revenue rests on two labs
Investor Steve Eisman argues AI's chain from Nvidia to Oracle depends on two cash-burning labs without moats, yet admits he cannot say whether this is a bubble.
Loading articles…
The complete subject coverage could not load. Published articles remain below.
Amazon is spending $220 billion and still leaving AI demand unmet
Amazon CEO Andy Jassy argues that staged capital commitments and sold-out capacity make a record AI buildout defensible, while conceding that the largest enterprise opportunity still depends on cheaper inference.
Nadella bets agents will dwarf cloud, if trust is earned
Microsoft's Satya Nadella argues long-running agents will expand the market far beyond cloud, but it demands new infrastructure, governance, model neutrality, and public trust.
AI made startups capital-bound: 20 people can now spend $1 billion
Martin Casado and Steven Sinofsky on AI's math wins and their unproven economic value, whether LLMs are another abstraction layer, and why capital now beats engineering.
Nadella saw search advertising early, then gave up on it
Satya Nadella and Ali Partovi trade confessions: how empathy was learned, why he stayed silent in a 2008 review, and why he regrets giving up on search advertising.
AWS redesigns the cloud for agents on $220 billion of capex
AWS's Matt Garman gets into designing the cloud for agents, rationing scarce GPUs between frontier labs and startups, and why $220 billion of capex is not a bubble.
Meta built Muse in 7 months with under 200 people
Alexandr Wang, who runs Meta Superintelligence Labs, covers the Muse personal agent, why Meta bet on consumer over coding, and how he rebuilt the lab as a flat, talent-dense team.