"I don't think a trillion is what we need to underwrite to at USV. I think if our biggest winners are in the single-digit billions, given our fund size and our ownership, we can produce really, really healthy returns for our partners." - Fred Wilson [00:00:07]
"The 900 really comes from that strategy and that model. We don't raise funds as vanity metrics... like 'I want to be a unicorn fund.'" - Mike Mignano [00:01:46]
Disclaimer: Orignal content owned by or sourced from third parties. It does not represent the views of 'Nuggets' platform or it's team. AI is used extensively across this platform including for summaries. Accuracy is not guaranteed, there can be mistakes. Any info or content on this platform is not a financial, legal, or investment advice. Do your own research. Refer for complete disclosures:- Terms of Use · Full Disclaimer
"I have a hard time imagining that we're just going to be delegating all of our computing away to agents. Yes, I think agents will do many things for us... but they're not going to watch Netflix for us. They're not going to watch YouTube for us." - Mike Mignano [00:13:14]
"I think we're entering a new phase where agents will be powering a lot of new experiences both for enterprises and consumers, and that's really, really exciting for us." - Mike Mignano [00:16:44]
"Trial behavior is at an absolute high, and engagement behavior is at an absolute low... We want to try them all, but we're going to stick with very few." - Rebecca Kaden [00:42:22]
"The best time to bet against an incumbent is when them doing what you want to do fundamentally threatens their current business." - Rebecca Kaden (quoting Brad Burnham) [01:06:23]
"Venture is based on an outlier mechanic. But we don't invest in companies that we don't see the opportunity ahead of us." - Fred Wilson [01:16:41]
Speakers & Credentials
Molly O'Shea - Host of Sourcery, tech interviewer, and venture capital ecosystem observer.
Fred Wilson - Co-Founder and Partner at Union Square Ventures (USV). Legendary venture capitalist with over three decades of early-stage tech investing experience (Twitter, Etsy, Coinbase, MongoDB).
Rebecca Kaden - Managing Partner at Union Square Ventures. Experienced thesis-driven investor focused on consumer platforms, healthcare transformation, and market restructuring.
Mike Mignano - General Partner at Union Square Ventures. Co-founder of Anchor (acquired by Spotify) and investor focusing on product-native startups, AI applications, and consumer software.
1. Executive Summary
Union Square Ventures (USV) has raised $900 million across its new fund vehicle cycle—marking the largest total capital raise in the firm's history while deliberately avoiding "vanity fund sizes" [00:00:00].
The increased fund size is driven by bottom-up market realities: Series A rounds now start at $10M–$30M+, startup go-to-market competition requires heavier upfront spend, and inference costs create substantial compute subsidies [00:09:31, 00:33:42].
USV maintains its core identity as a thesis-driven early-stage firm that targets 30–40 investments per core fund, with check sizes calibrated to secure high initial ownership without inflating fund scale past LP return dynamics [00:05:33, 00:32:24].
The firm concentrates its capital allocation across four main thematic pillars: "Obliterate, Don't Automate" AI applications, the open-source "Rebel Alliance" infrastructure, clean energy/nuclear power, and physical AI/robotics [00:29:52].
USV is aggressively expanding earlier in the funding stack via proprietary seed incubations (e.g., Supertake), venture partner-led originations, and founder-first seed bets to maintain ball control [00:07:31, 00:34:39].
In contrast to funds underwriting every company to a trillion-dollar valuation, USV’s model relies on generating extraordinary LP returns through $10B–$50B outcome outliers combined with strict check discipline [00:00:07, 01:19:31].
The partners highlight that traditional ARR metrics are currently distorted by upfront annual software subscriptions, asserting that true cohort usage retention is the ultimate predictor of long-term venture outcomes [00:41:01].
USV projects that incumbent legacy firms will fail to adopt pure AI offerings because doing so directly cannibalizes their existing human-capital-heavy business models [01:05:52].
[01:08:39] Security, Trust, Personal Agents, & Crypto Realities
[01:16:28] Realistic Valuation Underwriting: Do We Need Trillion-Dollar TAMs?
[01:21:18] What Keeps Partners Inspired: The Matrix-Level Founder
3. Detailed Thematic Summary
Macro Market Evolution & Sizing the $900M Fund Vehicle
USV’s new $900M capital raise represents its largest single fund cycle in over 20 years of operations, constructed from the bottom up rather than chasing arbitrary LP vanity thresholds [00:01:12, 00:01:46].
The structural shift reflects structural changes in software economics: while building a v1 prototype costs less than ever, the cost to scale, break through market noise, and fund GTM has exploded [00:09:24, 00:17:09].
Standard Series A valuations and check requirements have reset dramatically, rising from historical $3M checks in USV’s early days to minimum baseline checks of $10M–$30M today [00:33:30, 00:34:18].
Capital must also absorb heavy ongoing compute costs; startups are actively forced to subsidize user inference costs to gain rapid market share against free products like ChatGPT, Claude, and Gemini [00:09:31, 00:20:20].
LP expectations aligned seamlessly with the increased fund size because institutional limited partners recognize the exact same market expansion dynamics across their broader portfolios [00:29:16, 00:32:04].
The Four Core Investment Pillars
Obliterate, Don't Automate: Moving beyond standard B2B SaaS automation tools to back companies that completely destroy middleman gatekeepers and restructure multi-billion dollar verticals [00:08:48, 01:01:47].
The Rebel Alliance: Capitalizing on developer sentiment that rejects locked-in, closed-source foundation models (Anthropic, OpenAI) in favor of open-source/open-weight harnesses, custom routing, and local deployment stacks [00:57:46].
Energy & Infrastructure: Investing 20%–30% of the fund’s capital in abundant, programmable power generation (such as small modular nuclear reactors) as a highly levered baseline bet on scaling AI compute [00:19:07, 01:01:25].
Physical AI & Edge Computing: Allocating resources toward full-stack robotics companies building proprietary foundation models, spatial hardware sensors, and edge-device local execution [00:19:20, 01:11:28].
Early-Stage Origination: Seeds, Incubations, and "Ball Control"
To secure target ownership stakes without getting priced out of late seed/series A hype cycles, USV is targeting proprietary seed originations for 33%–50% of its early-stage deals [00:36:24].
The firm executes this through internal incubations built directly inside the partnership (e.g., Supertake, an AI agent that deploys themed equity investments via connected Robinhood/Coinbase accounts) [00:07:31, 00:35:17].
USV leverages senior Venture Partners (like Scott Belsky and Jared Hecht) to co-create ideas, recruit external operators, write the initial two funding checks, and maintain "full ball control" before exposing companies to external venture syndicates [00:35:26, 00:36:02].
Evaluation Paradigms: Why ARR is Flawed & What Cohorts Reveal
Annual Recurring Revenue (ARR) has turned into a misleading metric due to mandatory upfront annual subscription barriers that obscure real consumer sentiment [00:41:01, 00:41:31].
Contemporary consumer behavior shows record-high initial trial rates paired with record-low long-term engagement; users pay to test hype products but quickly churn [00:42:22].
Primary due diligence requires looking past superficial ARR growth curves and auditing daily/weekly cohort retention graphs to ensure the product builds habit-forming utility [00:37:53, 00:41:50].
In early-stage investing where multi-year retention data does not yet exist, USV relies on evaluating whether founders possess the narrative power required to win hyper-competitive hiring wars [00:38:59, 00:39:35].
The Incumbent Dilemma & The Future of Interfaces
Legacy services (law firms, tax preparers, asset managers) will fail to compete with AI native platforms because adopting zero-marginal-cost software destroys their human billable-hour revenue models [01:05:58, 01:06:38].
Publicly traded incumbents face immense earnings pressure that makes it impossible to intentionally burn existing revenue lines to deploy autonomous AI agents [01:07:40].
Contradicting viral Silicon Valley claims that text-only chatbots will replace graphic interfaces, USV argues that rich visual UIs remain essential for establishing user trust, visual hierarchy, and brand preference [00:12:40, 00:13:06].
The Reference Vault
4. Core Frameworks & Mental Models
Obliterate, Don't Automate [00:08:48, 01:01:47]: Originally formulated by Fred Wilson in 2014, this mental model rejects building B2B SaaS software that incrementally automates legacy workflows. Instead, it advocates for deploying technology that strips out middleman gatekeepers, restructures entire industry cost bases, and returns value directly to consumers.
The Rebel Alliance vs. Enclosed Stacks [00:57:46]: A framework dividing the AI ecosystem into centralized mega-labs (OpenAI, Anthropic, Google) and an open-source "Rebel Alliance" (open-weight models, custom routing, local execution). Developers default to open stacks to avoid platform risk, margin compression, and sudden API access termination.
The Incumbent’s Dilemma (Burn Your Own Boats) [01:06:23]: Quoting partner Brad Burnham, USV asserts that the optimal moment to bet against an incumbent is when the required technological shift fundamentally threatens their existing core profit engine. Established firms cannot burn their own boats without destroying shareholder value.
Single-Player to Multi-Player Network Transition [00:23:22]: AI products currently suffer from retention drops because they function primarily as single-player productivity utilities. Long-term defensibility requires converting single-player tools into multi-player networks that produce true network effects.
The Matrix-Level Founder [01:23:48]: A qualitative framework describing elite founders who evolve from early operators into visionary leaders capable of "seeing the Matrix." These founders combine intense narrative magnetism, world-class talent acquisition, and relentless iteration through long non-linear plateaus.
5. Anecdotes
The Evolution of Abridge (Healthcare AI Pivot) [00:44:20, 00:48:53]: USV backed Abridge in 2016 based on an initial consumer thesis: patients recording their own doctor appointments. The original consumer GTM failed to scale. Founder Shiv Rao recognized that doctors—not patients—wanted automated clinical notes. Pivoting to a physician-facing workflow built one of the strongest healthcare AI platforms in tech.
Duolingo’s Translation Origin Story [00:50:17]: When USV backed Luis von Ahn at Duolingo, the company's initial business model centered on crowdsourcing translation for foreign-language web documents rather than consumer language learning gamification. Bet on exceptional founders willing to abandon non-working models.
Coinbase’s Non-Linear Journey & Cold Storage [00:48:15, 01:13:31]: Coinbase’s growth trajectory was far from linear—experiencing flat multi-year stretches interrupted by massive crypto cycles in 2017 and 2019. Coinbase won consumer dominance by prioritizing compliance, security, and cold storage to protect users from scam-heavy environments.
Coinbase Engineering Dropping Claude for Open Source [01:00:02]: At USV’s CEO summit, Coinbase’s CTO shared that their engineering team shifted away from Anthropic’s Claude to build an in-house coding harness using Pi—an open-source framework built by USV portfolio company Arendelle.
Fred Wilson’s Crypto Security Experience [01:12:14]: Reflecting on personal security risks, Fred Wilson revealed that over 15 years of holding crypto assets, he was hacked four separate times due to early operational mistakes—yet retained over 99% of his total net crypto value, reinforcing the critical need for user-friendly security abstraction.
Zach Canter’s 7-Year Plateau at Steady [01:22:44]: Rebecca Kaden highlighted Steady founder Zach Canter, who persevered through seven difficult, non-linear years before product-market fit aligned over the last three years, demonstrating the endurance required in venture-backed companies.
6. References & Recommendations
Books
The Messy Middle by Scott Belsky [00:06:48]: Highly recommended by Mike Mignano as the definitive guide to navigating the hard, non-linear operational middle phase of building a startup.
Shoe Dog by Phil Knight [00:07:11]: Mentioned as a foundational business memoir on the USV library bookshelf.
Companies & Startups
Union Square Ventures (USV) [00:00:00]: Early-stage thesis-driven VC firm located in New York City.
XBloom [00:02:30]: Automated robotic pourover coffee machine showcased in the USV kitchen.
Board [00:03:00]: Tabletop digital gaming console portfolio company founded by Ben Upham.
Supertake [00:07:31]: Internal USV-incubated app that converts thematic ideas into executed stock/crypto portfolios via AI agents.
AI@GSB Dean’s Perspectives: Rick Osterloh, SVP at Google | 14 Oct 2026 | Stanford Graduate School of Business
"In most technology spaces when a new category is introduced it's typically additive and it's... usually coming from high value products... Will a new AI device come along that displaces the phone and totally replaces its usage? Most likel…
Current Series A Check Range
$10M to $30M
Target Series A check written per company under the new fund strategy