VCs deploying AI agents for deal flow as software moats collapse

Apr 15, 2026 · The GTMnow Podcast
🎧 PodShort 57 min squeezed to 2 AI SprinklerAS AI / ML
Episode artwork
Oren Hoffman
General Partner at Flex Capital, CEO at Incubate, Host of The Summation Podcast at Flex Capital
The GTMnow Podcast
57 min squeezed to 2
Full episode from The GTMnow Podcast
Quotable Moments

I am confident though that by the end of this year, by the end of 2026, most of our first meetings with companies will be agent-to-agent.

Oren HoffmanGeneral Partner at Flex Capital, CEO at Incubate, Host of The Summation Podcast
33:28

If they're not still good in three months, we will churn.

Oren HoffmanGeneral Partner at Flex Capital, CEO at Incubate, Host of The Summation Podcast
00:57
Key Insights
  • Most first meetings between venture capital firms and companies will transition to an agent-to-agent interaction by the end of 2026.
  • The current AI infrastructure build-out is fundamentally different from the dot-com bubble, being primarily funded by cash flow and equity, and driven by existing customer demand, rather than debt.
  • OpenAI's acquisition of TBPN (The Browser Company) was strategically about distribution and influencing key audiences like CEOs and investors, not merely about revenue or acquiring new broad audiences.
  • Software spend will double, yet many software companies will fail, and the Series A funding round has evolved into a 'slaughterhouse' due to intense competition and high failure rates.
  • In the future, software products must improve significantly (e.g., monthly updates) to retain customers, as traditional competitive 'moats' have largely disappeared due to intense market competition.
  • AI is predicted to cause a massive baby boom in the US by making family life easier and more affordable for affluent individuals, potentially by reducing the importance of college and easing household burdens.
  • The most painful mistake a venture capitalist can make is missing out on investing in a company that subsequently becomes very successful, outweighing the pain of making a bad investment.
  • Venture firms should critically assess why a deal is presented to them, as top companies typically secure funding early, and a late or widely circulated deal can signal underlying issues.
Metrics Mentioned
  • 500-600 AI analysts (The number of AI analysts on Oren Hoffman's team at Flex Capital.)
  • 53 investments in 2025 (Flex Capital's investment activity in a single year.)
  • Roughly 1 investment per week (Flex Capital's investment pace.)
  • $500,000 typically invested (Flex Capital's typical check size in a seed round.)
  • 2/3 equity to founders, 1/3 to Incubate (The equity split when Incubate starts companies with founders.)
  • ~$9.5 million raised (Funding secured for 2 out of 3 Incubate companies last year.)
  • 300+ companies (Oren Hoffman's personal angel investment portfolio.)
  • 200+ investments (Investments Oren Hoffman made with his own money before starting Flex Capital's fund.)
  • 40% down (Horizontal SaaS performance in public markets over the last 12 months.)
  • 2% down (Infrastructure performance in public markets over the last 12 months.)
  • 3% down (Vertical SaaS performance in public markets over the last 12 months.)
  • $20 billion in ARR (Reported annual recurring revenue for OpenAI and Anthropic.)
  • $0.37/share (Cisco's peak EPS.)
  • $4.06/share (Nvidia's current EPS.)
  • $3 million (Typical size of a seed round Flex Capital participates in.)
  • $100 million (Amount of money that typically wants to go into a $3M seed round, indicating high competition.)
  • 10 years later (The typical timeframe for venture capital investment outcomes.)
  • 5-10 breakout companies per year (Estimate of truly successful companies emerging annually.)
  • 20 breakout companies per 3-year fund (Total estimated breakout companies within a typical fund's lifecycle.)
  • 1 short vacation (A CEO of a billion-dollar revenue portfolio company has taken only one short vacation in 7-8 years.)
  • $200,000/year and up (Income level of 'quite wealthy people' in the US, defined as top 5% earners.)

RevBots.ai View:

  • AI Sprinkler stage evident in VC's use of 500 AI analysts for deal sourcing.
  • Distribution-first strategies (like OpenAI's) outperform pure product plays in ARM.
  • Monthly product updates required to retain customers mirrors ARM's velocity demands.
  • Tab Hopper founders face brutal Series A 'slaughterhouse' without AI leverage.
🎧Full Episode:The GTMnow Podcast →