“a lot of legal tech startups and uh public companies were hit hard. Thompson Reuters down 20%, Lexus Nexus, which is a database of case law, uh was down 15%”
Claude legal tools attack the legal-data moat; agentic layer eats the profit pool, terminal value plummets
“By the way, the numbers you report are dramatic understatement. We we've wiped out trillions of dollars in market cap. Figma's down 80% from the high”
Design tooling gets squeezed as vibe-coding collapses the UX/design/dev job functions into one
“It could be true that you're not going to replace CRM, but it can also be true that it's never going to trade at 30 times free cash flow again. And it's going…”
Gerstner's call is not that Salesforce dies but that its valuation re-rates permanently lower as the agentic layer captures the profit pool the app layer used to own. The stake is on the multiple, roughly halving from 30x to 17x free cash flow, not on revenue collapsing.
“Snowflake reaccelerating, Clickhouse reaccelerating. There are beneficiaries in the software space. Absolutely. Because it's all these AI tools rely on data…”
The counter-leg to Gerstner's app-layer short: data-platform companies reaccelerate because every agentic tool depends on the data and transformation that lives in their platforms, not in a thin layer on a CRUD database. Snowflake is the cleanest public name he stakes.
“One person being able to do three or four jobs, it just changes the nature of how profitable a company like Amazon, which is like my number one pick for like…”
Calacanis frames Amazon as the prime beneficiary of agentic consolidation: a giant labor-heavy operation where one worker now does three or four jobs, so margins expand without proportional headcount. He explicitly names it his top pick, not a passing example.
“I happen to think that I would take the over on the number of rate cuts that Warsh is going to give us this year. But I think the market is clearly a little…”
The market prices Warsh as a hawk: 'no cuts in 2026' sits at ~67c and one cut at ~18c. Gerstner argues anchored sub-consensus inflation plus an above-neutral restrictive rate means Warsh cuts more than the tape fears, so he takes the over on cut count.
“There are tons of retail investors and institutional investors like us who want to bet against that future, right? Then Elon's your guy and the combination of…”
Gerstner stakes that the SpaceX-xAI combination, merging the two biggest TAMs of AI and space, draws overwhelming retail and institutional demand into the IPO. He misspeaks 'against' and is corrected on-air to 'bet with that vision,' counting himself among the buyers.
“I do think that we will see compute efficiency scale by probably on the order of 70 to 100 X over the next few years. Meaning electricity efficiency per token…”
Energy is the binding AI constraint; chip+model architecture drives 70-100x efficiency per token