All-In

Ben Horowitz

CoreWeave’s 953-basis-point CDS implies 16% annual default risk; fixed-price contracts and $35 billion more to raise leave Chamath’s team estimating a $5 billion break-even gap.Sacks sees demand holding, but financing costs remain a risk; Chamath urges Anthropic to raise at least $200 billion for compute, with SemiAnalysis estimating it and OpenAI will take 60% of next year’s global additions.

Martin Shkreli

Shkreli limits quantum’s edge to factoring and elliptic-curve logarithms; current machines lack quantum RAM.He calls industry spending speculative and expects analog optical computing to challenge Nvidia’s H100, B100, and B200 within 20 years.His pharmaceutical thesis favors mispriced discarded molecules; he calls AI drug discovery within five years “nonsense” and expects no AI-related drug in the next 10 years.

Chamath Palihapitiya × Jason Calacanis × David Sacks × David Friedberg × Jake Paul × Drew Taggart × Alex Pall

Jake Paul is turning owned attention into distribution and talent infrastructure, citing 138 million Netflix viewers for the Mike Tyson event and 400 MVP fighters.His $100 million Anti Fund and The Chainsmokers’ fund target venture economics, but later tranches priced at two or three times earlier valuations without operating improvement leave realized DPI—not TVPI—as the test of Chamath’s bubble warning.

Luca Ferrari

Bending Spoons converts acquired product-market fit into a repeatable model, using 50+ proprietary tools and an approximately 800-person shared talent base to replace technology foundations.Historically consistent unlevered returns above 25% and debt costing about 9%, fully hedged and maturing in 2031, support the model, while limited customer synergies, competition and replication risk remain watchpoints.

Jason Calacanis × David Sacks × David Friedberg × Brad Gerstner

Google’s infrastructure-first capital shift is coinciding with senior AI departures, while Sacks sees Anthropic and OpenAI forming a premium frontier duopoly as cheaper models pressure weights toward commodity pricing.SpaceX reported $7.8B revenue, up 92% YoY, and Starlink’s cash generation could fund expansion, but the proposed $300B-scale AI buildout leaves financing, spot pricing, and demand sensitivity as key risks.

Mark Cuban × Chamath Palihapitiya × Jason Calacanis × David Sacks × David Friedberg

The AI bubble’s likely casualties are VCs, funds, and PE firms concentrated in peak-priced private rounds, while infrastructure spending depends on whether token economics and video demand justify today’s buildout.Enterprise implementation remains the bottleneck as agents drift and require forward-deployed talent, but falling build costs are opening bespoke software and IPO-funded acquisition strategies for smaller AI companies.

Bill Maris

Bill Maris argues that venture funds below $750 million structurally outperform megafunds, citing 4.76x versus 2.42x top-decile DPI and the exit-value hurdle facing a $7 billion fund.Google’s war chest could compress OpenAI and Anthropic if Gemini offers comparable products at 80% lower token prices, while AI’s near-term opportunity may lie beneath models in memory, controllers, GPUs and other enabling layers.

Brad Gerstner × Gavin Baker × Kelly Rodriques × Chamath Palihapitiya × Jason Calacanis × David Sacks × David Friedberg

Private-company secondaries have become a third exit market, with 2025 volume roughly twice the 2021 peak and activity equal to 31% of primary venture activity.Forge’s Schwab partnership could distribute private equity to millions of investors through regulated products, while longer private lifecycles, double-fee SPVs and valuation compression make governance, pricing and liquidity worth monitoring.

Andrew Feldman × Will Marshall × Brad Gerstner × Chamath Palihapitiya × Jason Calacanis × David Sacks × David Friedberg

The IPO pendulum is shifting toward companies listing at $1 billion to $5 billion, with Planet creating roughly 90% of its post-SPAC value in years three and four.Cerebras’s $18.50 IPO followed a difficult 9.5 years, while its chip targets 15–18-times faster OpenAI workloads; orbital data centers offer a potential two-to-three-year cost catalyst, but distributed clustering remains unresolved.

Thomas Laffont

AI has made the unicorn economy healthier but sharply more concentrated: funding per unicorn is up 5x since 2021, and a private “Magnificent Eight” approaches $4 trillion ahead of a possible liquidity event.OpenAI and Anthropic could exceed AWS by year-end and potentially all of Microsoft by 2028, while SpaceX’s recurring constellation revenue supports a platform thesis; public-market scrutiny and a possible price war remain unresolved.