Sequoia & Union Square Returns, the Post-AI Labor Market, and the Return of SF? | E1907

Episode Summary

Executive Summary: The panel examines leaked VC returns, arguing that venture outcomes are heavily driven by vintage timing, fund discipline, and power-law winners like Union Square Ventures. They then discuss how Elon-style lean operations and AI/offshoring are reshaping startup hiring, and close by debating regional migration back to San Francisco plus new investment ideas across defense, energy, AI, and manufacturing.

Main Topics: Leaked VC returns and vintage effects (Priority: 5/5): The hosts react to UTIMCO’s disclosed venture returns and use Union Square Ventures as a case study for how older funds can compound dramatically while newer vintages remain too early to judge. Power law, fund size, and LP expectations (Priority: 5/5): The discussion emphasizes extreme dispersion in VC, the importance of manager selection, and why LPs prefer consistent exit policies over ad hoc decision-making. Startup staffing, efficiency, and the 'Elon model' (Priority: 5/5): The group argues that Elon Musk’s Twitter/X restructuring normalized operating with far fewer employees and inspired founders to pursue leaner teams and more rigorous role evaluation. AI, offshoring, and the startup employment market (Priority: 4/5): The panel debates whether layoffs and hiring softness reflect AI disruption or more immediate pressure from offshoring, with broad agreement that companies will increasingly automate, delegate, and deprecate work. One-person unicorns and AI-induced demand (Priority: 4/5): They speculate about when a one-person unicorn may emerge, with one view that AI will lower the cost of intelligence but also expand total demand for labor and new company types. Geography, San Francisco, and network effects (Priority: 3/5): The hosts revisit Miami vs. San Francisco migration and conclude that network effects remain strongest in the Bay Area, even as Austin, New York, Miami, LA, and Texas continue growing. Latest investments in deep tech and software (Priority: 3/5): Each panelist shares recent bets spanning nuclear microreactors, defense energy, cybersecurity, manufacturing automation, AI-assisted engineering, consumer hardware, and sports coaching tools.

Key Arguments: VC returns must be judged over long horizons because the J-curve makes newer funds look weak before exits materialize. Older vintages launched near crises often outperform, suggesting that bear markets can be attractive entry points for venture investing. Union Square Ventures is notable not just for one outlier but for multiple winners across Twitter, Coinbase, Etsy, Tumblr, Mongo, Zynga, and LendingClub. LPs value clear and consistent exit policies more than maximizing every single position because surprises and style drift are costly. Elon’s Twitter/X restructuring showed that a company can ship more with a fraction of the staff, changing founder expectations about team size. The near-term labor threat is likely offshoring and efficiency tooling, while AI becomes more disruptive over a longer horizon. AI lowers the marginal cost of intelligence, which may create new roles, new companies, and potentially new funds rather than simply shrinking the market. San Francisco’s dominance comes from dense network effects and U.S. institutional advantages, not just geography or weather. The U.S. venture ecosystem remains uniquely powerful because of university-tech transfer, risk tolerance, and a culture that rewards failure and experimentation.

Data Points: UTIMCO AUM: $65 billion - University of Texas endowment that was compelled to disclose VC returns USV return for UTIMCO: 9.14x cash-on-cash - Across nine funds, with caveat that older funds still have room to appreciate VC market size in the 1990s: about $30 billion - Approximate size of the venture market thirty years ago Global VC market size today: $300 billion to $400 billion - Approximate current size discussed by Guy Typical VC return range: 1.8x to 3.2x - Broad return band over long intervals for many venture funds Top-vs-bottom dispersion: 10x to 60x - Illustrating the variance between top and bottom venture managers Twitter staffing reduction: 15% of the people - Used as an example of Elon Musk’s lean operating model January 2024 layoffs vs quits: Layoffs exceeded quits for the first time since February 2023 - A sign of weakness in the startup employment market OpenPhone discount: 20% off first six months - Podcast sponsorship offer Squarespace discount: 10% off first purchase - Podcast sponsorship offer Coda startup credit: $1,000 startup credit - Podcast sponsorship offer Add/Delegate/Deprecate framework: every quarter - Jason’s internal operating cadence for reviewing work and efficiency Threshold for selling positions: 3x mark - Guy’s policy for exiting positions once entry cost versus market cost crosses this level Seed-stage liquidation approach: 10% sold two or three times on the way up - Jason’s stated partial-exit philosophy at seed stage Twitter 2012 USV fund outcome: 22x - Jason inferred this fund likely benefited from Twitter and other winners Microreactor size: 300 kilowatts - Antares’ proposed diesel-generator-sized nuclear microreactor Offshoring threat model: fraction of cost - Eric’s view that startups will replace some U.S. hires with offshore teams before AI fully substitutes them

Pivotal Quotes: "that framework, I think, that Elon came up with was elegantly simple, yet very complex" — Jason Calacanis: Describing the Twitter/X staffing framework of evaluating employees by whether they are exceptional and/or essential "The one thing that I think investors hate, not maybe as much as losing money, but right up there are surprises, right? Inconsistencies." — Guy Perelmuter: Explaining why LPs prefer clear, repeatable exit policies from venture managers "what we’re witnessing right now is the creation of very efficient intelligence, right? Inexpensive and efficient intelligence." — Guy Perelmuter: Arguing that AI will expand the market by making intelligence cheaper and more widely available

Implications: Listeners should expect more disciplined, leaner startups, greater scrutiny of VC performance, and a continued shift toward automation/offshoring. AI may reduce headcount needs, but it could also spawn new company types, roles, and capital strategies.