E118: AI FOMO frenzy, macro update, Fox vs Dominion, US vs China & more with Brad Gerstner

The AI boom was already crowded: roughly 500 generative-AI startups had raised more than $11 billion, excluding Microsoft’s OpenAI investment. The investors broadly agreed AI was a major platform shift, but argued foundation models may commoditize while scarce data and specialized silicon capture more durable value.

OpenAI’s 90% API price cut pointed toward a platform strategy rather than simply selling an expensive model. Cheap access lets products such as Notion add AI without building models themselves, potentially making OpenAI a small but pervasive “tax” on thousands of applications.

Venture investors said institutional LPs were concentrating money into established firms rather than abandoning venture altogether. Brad Gerstner argued first- and second-time funds without demonstrated distributions would struggle, while large LPs would keep long-term allocations but narrow their managers sharply.

Early-stage investing had become materially more disciplined after the 2021–22 crash. Jason Calacanis described $5–10 million startup valuations, six-week fundraising processes, milestone-based rounds and founders again presenting customer-acquisition models while scrutinizing salaries, hiring and capital deployment.

The cleanest fix for an overvalued startup, the group argued, is often to reset the valuation and restore employee equity rather than engineer complicated financing. Chamath Palihapitiya described one troubled company where they halved the valuation and reissued options because existing employees had become economically underwater.

Silicon Valley’s efficiency push followed extraordinary headcount expansion: Salesforce grew from roughly 19,000 employees in 2015 to 80,000 seven years later. Benioff subsequently made profitability his top strategy, while the discussion cautioned that Twitter-style mass cuts could break larger public companies if applied too abruptly.

Stock-based compensation emerged as one of the episode’s sharpest criticisms of tech-company accounting. Booking.com reportedly diluted shareholders about 5% over 15 years versus roughly 25% at Salesforce and Expedia; the argument was that excluding stock compensation from adjusted earnings hides a genuine shareholder expense that buybacks later consume cash to offset.

The macro picture had shifted from expectations of falling rates toward “higher for longer,” even as markets remained unexpectedly resilient. The two-year Treasury yield had moved from roughly 4.1% to 4.9% in a month, while Gerstner stressed that short-term macro forecasting remained unreliable enough that investors should prioritize businesses able to grow through different rate environments.