The Twenty Minute VCWhy VC Today is Worse than 2021
CHAPTERS
Benchmark adds Everett Randle: partner moves and firm resilience
The conversation opens with Everett Randle joining Benchmark and what it signals about Benchmark’s culture, portfolio strength, and recruiting playbook. The hosts frame the move as part of a broader “VC shuffle” where ambitious investors move quickly between top firms.
Is VC still the best-paid job in tech? Talent markets and compensation reality
They debate whether venture is still the most lucrative career path compared to top AI engineering roles. Rory argues liquid, mega-equity packages at companies like Meta outcompete VC economics in the near-term, even if venture can win over decades.
Carry is slow: delayed distributions, privatization, and the ‘get rich slow’ problem
The group digs into how long it takes to see meaningful carry, especially with companies staying private longer. They discuss how paper gains don’t translate into distributions and how holding winners can delay liquidity further.
Revolut at $75B: private markets replacing public markets
Revolut’s $75B valuation becomes a case study in late-stage private funding substituting for IPOs. They explore what fundamentals justify the round and how private pricing forces everyone to underwrite massive future TAM expansion.
The TAM myth vs TAM exhaustion: why market size anxiety is rising
Jason challenges the classic belief that great founders can always expand their TAM, arguing TAM exhaustion is appearing across portfolios. Rory agrees on exhaustion risk but argues the best companies start in a niche and expand into adjacent empty space.
Why Spotify won: licensing strategy, geography, and execution vs ‘great man theory’
They briefly unpack Spotify’s early advantage: building in European markets with less aggressive label scrutiny, securing better licensing, and reaching critical mass before taking on the US. The segment reinforces their theme that outcomes combine founder skill, market structure, and timing luck.
Vertical AI hype and the ‘Covid mistake’: everyone is suddenly “in-market”
Jason argues investors are repeating 2020-style extrapolation errors: AI is putting every CIO and function “in-market” at once, inflating near-term demand signals. Rory calls it a huge point—compressed buying cycles can mislead growth forecasting and valuation setting.
How founders should react: win fast, model onboarding costs, and know when to sell
They pivot from investor theory to founder strategy: if the market window is compressed, leaders must move quickly to lock in customers before decisions are made. They also emphasize overlooked implementation and change-management costs, and Jason introduces a sell/hold rule based on TAM acceleration.
Vertical SaaS ‘truth’ in the AI era: Toast as the benchmark and the valuation trap
They debate whether vertical SaaS can still generate venture-scale outcomes when exits effectively require far larger revenue. Jason argues many verticals are overfunded relative to realistic ceilings; Rory agrees the businesses can be valuable but warns against paying prices near total TAM.
OpenAI’s cloud shift: Oracle vs Microsoft and who carries the balance-sheet risk
They interpret reports that OpenAI may spend more with Oracle than Microsoft as Microsoft refusing economically irrational capex. The hosts argue OpenAI is exceptionally good at offloading infrastructure risk to partners while preserving its own upside.
Poolside builds a data center: AI’s rising capital intensity and compute scarcity
Poolside’s plan to build a massive AI data center becomes evidence that competing at the model/application layer may now require owning infrastructure. Rory frames it as a terrifying escalation from software economics to fixed-asset economics, likely driven by inability to secure enough compute capacity.
Boom, bust, and ‘temporal diversification’: surviving corrections in AI investing
They discuss what a real AI/data-center bust would look like (mild glut, falling prices, halted capex), while agreeing the timing is unknowable. Rory argues the key is balancing aggressiveness with survivability, but notes the industry is compressing fund cycles and abandoning temporal diversification.
Content boundaries and privacy: OpenAI allows erotica and the moderation hot seat
The hosts close with a discussion on OpenAI permitting erotica, using it as a proxy for broader boundary-pushing in AI outputs. They highlight how AI differs from social platforms because the model generates content directly, increasing accountability and making moderation far harder.
Rapid-fire bets: Replit to $1B ARR, Deel vs Rippling, and what ‘early’ means now
In a game-format finale, they debate whether Replit can hit $1B ARR quickly and how AI tools change early-stage evaluation. They also compare Deel and Rippling via TAM defensibility and pain intensity, ending on the paradox that even $1B ARR can feel ‘early’ under today’s venture expectations.
