LIQUIDITY: Major challenger to OpenAI, xAI, closing on $6B in funding | E1944

Episode Summary

Executive Summary: The episode blends Knicks fandom with a deep venture-capital discussion. The panel argues that AI is becoming commoditized, so durable winners will rely on proprietary data, scale, talent, and capital intensity. They also examine Pitchbook’s case for emerging managers, concluding that access, fund size, GP commitment, and hunger matter more than simple vintage labels. The show closes with three personal investment picks each.

Main Topics: Knicks playoff obsession and sports banter (Priority: 3/5): The episode opens with extended conversation about the Knicks' playoff run, the emotional toll on fans, and playful betting around the Knicks-Pacers series. XAI, OpenAI competition, and the AI capital war (Priority: 5/5): A major segment explores Elon Musk's XAI raise, the commoditization of LLMs, and why unique data, scale, and war-chest size may determine the winners. Open source vs proprietary AI and enterprise adoption (Priority: 5/5): The panel debates whether open source or closed systems will dominate AI, drawing analogies to search, TikTok, Facebook, and enterprise software adoption patterns. Emerging managers vs established VC firms (Priority: 5/5): They discuss Pitchbook’s report on emerging managers outperforming established firms, but argue the finding is incomplete without considering survivorship bias, fund size, and power-law investing. Incentives, GP commitment, and fund structure (Priority: 4/5): The conversation turns to alignment: personal capital commitments, carry, preferred returns, and how bigger funds can distort incentives or reduce hunger. SVB Capital sale and ecosystem churn (Priority: 4/5): They analyze the SVB Capital fire sale and implications for talent retention, LP access, continuation vehicles, and the broader post-bubble VC cleanup. Latest investments and founder/company updates (Priority: 3/5): Each speaker shares recent investments across fintech, permitting, advisory software, podcast search, ML-backed manager funds, and retail enablement tools.

Key Arguments: AI models are becoming commoditized, so differentiation will increasingly come from proprietary data, distribution, and scale rather than model quality alone. XAI is likely a capital-intensive, winner-take-most battle that may be more suited to giant strategic backers or sovereign-like capital than traditional VC. Open source often wins in infrastructure and enterprise tooling, while closed systems can win in consumer products; AI may split along those lines. Pitchbook’s emerging-manager outperformance data is interesting but incomplete because it ignores survivorship bias and the fact that only a small subset of emerging managers are truly exceptional. LPs should think in terms of backing companies through GPs, and because venture is power-law driven, more high-quality exposure can improve outcomes. Smaller, focused funds may outperform because they preserve GP discipline, keep fund managers hungry, and improve decision quality and founder support. GPs with meaningful personal capital at risk are better aligned and often more helpful to founders during downturns. Some of the post-2021 venture ecosystem was inflated by tourists and easy capital; the downturn should flush out zombie funds and restore discipline.

Data Points: XAI raise size: $6 billion - Reported size of Elon Musk's AI competitor round discussed on the panel XAI valuation: $18 billion - Valuation referenced by David when describing Sequoia’s investment Sequoia return target: 3x to 5x - Described as the more typical growth-equity-style outcome versus venture-style 100x LP commitment / GP commitment: ~10% - Jason said he expects to put about 10% of his $50M fund into the vehicle, including fees/salary LP portfolio construction: 10 to 15 funds per year - Joshua said his LP strategy is to back that many funds annually Fund evaluation volume: 300 to 400 people contact me - Jason said this is roughly how many fund managers contact him Team screening rate: 10% - Jason said his team meets with about 10% of the fund managers who contact him Emerging manager allocation count: 1 new manager per year - Jason said he adds about one new manager per year now Venture application conversion: 200 applications to 1 investment - Jason said his firm is now roughly at this ratio YC conversion benchmark: 1% - Referenced as Y Combinator's approximate investment rate for comparison Current firm conversion rate: 0.5% - Jason said his firm is now investing at about this rate Podcast platform inventory: 2,000 episodes of This Week in Startups and 177 All-In episodes - Jason mentioned the scale of his audio archive in the podcast-search discussion Sparkplug customers: 2,500 retailers and 500 brands - Jason cited Sparkplug's current footprint Liquidity Summit capacity: 125 people - Jason described the annual event size Liquidity Summit speaker count: About 25 speakers - Jason described the event’s structure Liquidity Summit attendee mix: About 60 GPs, 30 to 40 LPs - Jason detailed the expected audience composition SVB attrition estimate: 10% to 20% of managers - Jason estimated manager churn at the old SVB network Potential additional attrition: Another 20% to go - Jason suggested more managers may become zombie funds or wash out

Pivotal Quotes: "deal flow is destiny" — Jason Calacanis: Used to explain why LPs and GPs should maximize high-quality meetings and opportunities "No conflict, no interest" — Jason Calacanis: Cited as a Michael Moritz lesson about how overlapping relationships can create informational value "if you hate your life, you know, you're never going to be successful" — Donald Stalter: On why fund managers must genuinely enjoy venture investing to perform well and support founders

Implications: The discussion suggests AI winners will be capital-heavy and data-rich, while VC returns may increasingly favor disciplined, aligned, smaller managers. Expect continued ecosystem cleanup, more scrutiny of incentives, and stronger demand for founder-friendly, high-conviction capital.