Search for a command to run...

Timestamps are as accurate as they can be but may be slightly off. We encourage you to listen to the full context.
In this wide-ranging conversation, venture capitalists Jason Lemkin, Rory O'Driscoll, and Harry Stebbings dive deep into the current state of AI investing and venture capital. The discussion covers Benchmark's newest partner hire, Revolut's massive $3B funding round at a $75B valuation, and the increasing capital intensity of AI infrastructure. (30:00)
Jason is a prominent venture capitalist and founder of SaaStr, one of the world's largest SaaS communities. He's known for his expertise in B2B software investing and has been a leading voice in the SaaS industry for over a decade.
Rory is a veteran venture capitalist with extensive experience in enterprise software and B2B investing. He brings decades of market cycle experience and has witnessed multiple boom-bust cycles in the technology sector.
Harry is the host of 20VC and founder of 20VC, a prominent venture capital podcast and investment platform. He's known for his incisive interviews with leading investors and entrepreneurs in the startup ecosystem.
Jason draws a parallel between the current AI market and the COVID-era software boom, noting that every company is currently evaluating AI tools simultaneously rather than the typical 5% annual market penetration. (21:01) This creates an artificial sense of massive market opportunity that may not sustain long-term. The key insight is that businesses won't be purchasing new AI tools every year - they'll make a decision, implement it, and stick with it for several years due to the high soft costs of onboarding and business process changes.
With companies now expected to reach $1B+ in revenue before going public, Total Addressable Market limitations become critical much earlier in a company's lifecycle. (13:00) Rory emphasizes that when exit bars are set at billion-dollar revenue levels, many vertical markets simply aren't large enough to support the growth trajectories required. This means investors must be much more selective about market size from day one, rather than hoping founders will expand their TAM over time.
The discussion of Poolside's decision to build their own 2-gigawatt data center illustrates how AI competition now requires infrastructure-level capital commitments. (45:00) As Rory notes, smart companies are concluding they need to own their compute infrastructure to compete effectively, transforming what appeared to be a software business into a fixed-asset intensive operation requiring billions in upfront investment.
Given that everyone is in market simultaneously for AI tools, companies that capture market share now will have significant advantages when the buying frenzy subsides. (27:02) Jason argues that businesses exhaust themselves with vendor switches and business process changes, meaning the current window represents a once-in-a-cycle opportunity to establish market dominance. Companies that don't win significant market share in the next 1-2 years may find themselves permanently disadvantaged.
The podcast highlights that venture capital returns over the past five years have underperformed public markets while requiring significantly more operational complexity. (61:38) Rory points out that LPs need 300-400 basis points above public market returns minimum to justify the illiquidity, manager selection costs, and cognitive overhead of venture investing. This creates pressure on the entire venture ecosystem to deliver outsized returns or risk capital flight.