Skip to content
All-In PodcastAll-In Podcast

Why the datacenter backlash could cap America's AI lead

Ratepayer groups, doomer activists, and local politicians are blocking permits; Allbirds' AI-pivot stock surge shows how scarce compute has become.

Jason CalacanishostChamath PalihapitiyahostTravis KalanickguestDavid Sackshost
Apr 16, 20261h 30mWatch on YouTube ↗

At a glance

WHAT IT’S REALLY ABOUT

All-In debates NYC taxes, OpenAI vs Anthropic, datacenters, markets, politics.

  1. The panel criticizes New York City’s proposed pied-à-terre tax as targeting the most “elastic” buyers and likely reducing high-end demand, development incentives, and broader city spending.
  2. They frame OpenAI as facing a strategic identity crisis—consumer dominance vs enterprise/coding focus—while arguing Anthropic’s faster growth and release cadence could drive a lasting lead.
  3. A major theme is “datacenter wars”: compute, power, permitting, and local backlash are portrayed as the real bottlenecks that can cap AI lab growth regardless of product demand.
  4. They interpret meme-like market behavior (e.g., Allbirds’ AI pivot stock spike) as both bubble signaling and a rational response to compute scarcity and infrastructure value.
  5. The show alleges coordinated political “machine” tactics behind Eric Swalwell’s exit and debates why markets remain strong despite Iran-related conflict, emphasizing expectation management and valuation risk.

IDEAS WORTH REMEMBERING

5 ideas

Targeting second-home buyers may backfire on NYC housing and development.

Sacks argues pied-à-terre owners are the most price-sensitive segment because they can buy elsewhere, so an annual tax could reduce demand, impair project underwriting, and slow new construction rather than improve affordability.

Housing affordability improves most reliably through supply expansion, not punitive taxes.

They cite Austin as a counterexample where allowing new builds coincided with rent declines despite net in-migration, contrasting with NIMBY-heavy “blue city” constraints.

In frontier AI, growth rate and product cadence can outweigh headline brand dominance.

Kalanick emphasizes “growth is king” and network effects from scale; Friedberg and Sacks point to Anthropic’s rapid release cadence and enterprise traction as a powerful flywheel.

Enterprise (especially coding) monetizes better than consumer subscriptions.

Sacks argues consumers prefer flat $20 “all you can eat,” while businesses pay metered token usage like electricity, enabling revenue to scale faster for coding-centric vendors.

Compute access is becoming the strategic choke point for AI labs.

Chamath and Sacks argue hyperscalers control a majority of compute, which can throttle frontier labs; labs may need vertically integrated infrastructure (land/power/shell) to avoid “Friendster-style” reliability constraints.

WORDS WORTH SAVING

5 quotes

They’re targeting the most elastic part of the market... which will crash the whole market.

— David Sacks

Growth is the whole damn thing.

— Travis Kalanick

Anthropic was very focused on enterprise, specifically coding... Consumers have a lower willingness to pay.

— David Sacks

We are absolutely compute constrained. Massively.

— Chamath Palihapitiya

The data center... is the temple of the wealthy... the physical manifestation that people want to attack and destroy.

— David Friedberg

NYC pied-à-terre tax and high-end housing demandRule of law, capital flight, and development incentivesOpenAI vs Anthropic: valuation, growth rates, and enterprise pivotCoding models, token economics, and enterprise willingness to payCompute supply chain: GPUs, power, permitting, and NIMBY backlashPrediction markets and hype/bubble signals (Allbirds pivot)California Democratic machine politics (Swalwell)Market valuation indicators: Shiller P/E, Buffett indicator, dispersionAI adoption realities: agent “slop,” change management, ROI timingGame segment: “The Price Is Wrong” overvalued startups

High quality AI-generated summary created from speaker-labeled transcript.

Get more out of YouTube videos.

High quality summaries for YouTube videos. Accurate transcripts to search & find moments. Powered by ChatGPT & Claude AI.

Add to Chrome