Interview, Fireside Chat
The AI Boom Will Create Enormous Roadkill: Who Wins & Loses? | David Frankel
Market Structure and Fund Performance
- Venture capital is becoming a "pyramid" where missing the top 5 companies (potential $3T valuations) makes returning a fund significantly harder.
- Median performance of the top 500 companies created in the last 25 years is a $2.6 billion valuation, which is sufficient to return a fund if an investor holds 5% of the winner.
- $50M–$100M "middle-market" funds are described as the worst performing, being "too big to write collaborative $100k–$250k checks" and "too small to lead $8M–$10M seed rounds."
- Over 90% of companies founded during the current AI wave will likely fail ("roadkill"), though survivors like OpenAI, Anthropic, and SpaceX represent the new Meta/Google era.
- Less than 100 companies have sustainably reached a $10 billion valuation in the last 25 years.
- Seed investing remains viable due to the ability to wait for the right founder ("alchemy") rather than chasing every momentum deal.
- Founder Collective avoids leading Series A rounds to prevent "preemptive" behavior and maintains a uniform treatment of all investors.
- Founder Collective's largest LP is the General Partner (GP) itself, ensuring alignment with Limited Partners (LPs) focused on DPI (Distributed Profit Impact) rather than just TVPI (Total Value to Paid-In).
Investment Strategy and Valuation Dynamics
- Founder Collective views "billion-dollar valuations as the new Series A," noting companies like McCore ($20B), Cognition ($26B), and Cursor ($60B) exiting early.
- The firm has historically rejected deals based on ownership percentages (e.g., 1–2% stakes in 11 Labs, Granola, Star Cloud, Fractile), acknowledging these were "hundreds of millions of lost returns."
- Founder Collective rarely leads follow-on rounds; they participate as insurance policies for founders ($500k–$1M checks) to ensure brand validation and distribution.
- The firm believes "price matters less than ever" regarding being in the "true winners," though uncapped notes at the seed stage are financially disadvantageous compared to fixed-price deals.
- The firm is "intoxicated" by the opportunity to identify founders with the potential to create $5T companies (e.g., SpaceX, Tesla, Meta, Nvidia, Palantir, Suno, Shield AI).
- Founder Collective's strategy has shifted toward deeply engineering-specific founders emerging from labs like DeepMind and Gemini, valuing "psychographic focus" over pure industry experience.
- The firm argues that "SaaS Apocalypse" valuations may be overdone for mission-critical, embedded software (e.g., Olo, Palantir), while leaving room for "orphaned" companies that can be acquired.
- Founder Collective has never sold a share of SeatGeek despite a 14+ year holding, treating the investment as a "religion" despite the lack of a liquidity event.
- The firm prefers to sell 20% of a position early to return 25% of the fund's DPI, prioritizing cash velocity over holding for a potential 2x exit.
The AI Landscape and Geopolitics
- The current AI wave is characterized by rapid momentum, with companies like Suno reaching $5B valuations in very short timeframes.
- Founder Collective anticipates another "dot-com crash" is inevitable, though the timing is unknown.
- China is identified as the second AI superpower, with "open models" potentially disrupting incumbents like OpenAI and Anthropic due to less restrictive regulation and faster deployment speeds.
- Photonic computing is predicted to be the next major disruptor to Nvidia, potentially solving energy constraints in data centers.
- The U.S. government's heavy reliance on non-dilutive funding (e.g., Tesla under the Biden administration) is seen as a net positive for tech business, despite political narratives.
- AI is expected to increase productivity rather than cause mass unemployment, with a "binary" divide between those who can use the tools effectively and those who cannot.
- Younger generations (22-year-olds) possess a "mental plasticity" regarding AI tools that older workers (45+) lack, creating a competitive advantage for the youth.
- Services and industries requiring high-trust relationships (e.g., $100B litigation, medical audits) will retain human interfaces despite AI automation of "grunt work."
- The "Nepo Baby" label is redefined by Founder Collective to include founders with deep historical vertical knowledge (e.g., TJ Parker in pharmacy, Evan in HVAC), rather than just wealth.
Founder Dynamics and Portfolio Specifics
- Founder Collective prioritizes the "CEO/CTO alchemy," specifically seeking a CEO who is a strong salesperson and a CTO who acts as a "Swiss Army Knife" or "magician."
- The firm rarely invests in a company where the CEO is exceptional but the co-founder is not, though they acknowledge they have regretted saying "no" to such dynamics.
- Second-time founders who had a "life-changing" but not "home run" first exit are often "hungrier" and more focused than those who had a massive initial success.
- Founder Collective admits to being "less patient now" due to higher service standards, but emphasizes kindness and presence as the antidote to this trait.
- The firm views the current "SaaS Pocalypse" as a buying opportunity for top-tier stocks that have lost market cap but remain deeply embedded in enterprise workflows.
- Secondary markets have become highly liquid for top-tier companies, with some selling at premiums to the last round price due to high momentum.
- Founder Collective has invested in "defense" companies like Shield AI, despite it being a controversial sector for some partners.
- The firm's "frameworks" for investment decisions have saved them from "orphaned" companies but also caused them to miss outliers like Klaviyo.
Future Outlook and Personal Reflections
- Founder Collective believes the "risk premium" for seed-stage investing is no longer overstated, having narrowed the gap between early-stage value and later-stage perception.
- The firm is more excited about health discoveries regarding chronic conditions (e.g., MS, chemo) enabled by AI than by autonomous driving or consumer AI.
- The firm anticipates that in 5–10 years, humans will no longer need to drive cars, marking a generational shift similar to the transition from landlines to mobile.
- Founder Collective sees the "SaaS" model evolving into "Applied AI" and "Physical AI," where commoditized hardware (drones, cameras) is enhanced by AI.
- The firm has observed that "momentum" deals (e.g., Suno) are easier to exit quickly than "hardware" deals (e.g., Whoop), which require longer capital cycles.
- Founder Collective notes that "TVPI" (Total Value to Paid-In) is regaining popularity among LPs, though "DPI" remains the ultimate measure of success.
- The firm has not yet raised a larger fund because the "arbitrage" of seed-stage risk premiums has disappeared, making larger funds economically less attractive unless a value opportunity arises in Series A/B.
- Founder Collective emphasizes that "kindness" and being "present" (no phones at dinner, in the bedroom) are the most critical factors for a successful marriage and leadership style.