Podcast Summary: This Week in Startups – “The best founders find a way to make it happen”: VC Roundtable with Bryan Kim and David Clark | E2222

Date: December 11, 2025
Host: Jason Calacanis (stepping into a commentator role, moderated by Alex)
Guests:

• Bryan Kim (Partner, Andreessen Horowitz)
• David Clark (Chief Investment Officer, VenCap International)

Episode Overview

This episode is a dynamic venture capital roundtable focusing on the current state of AI investing, startup fundraising environment, company exits, and the ever-evolving competitive landscape of tech and markets. Jason Calacanis, Bryan Kim, and David Clark explore critical issues such as whether AI is in a bubble, how valuations are set in this new era, what makes a defensible business, the emergence of new business models, and lessons from founder resilience. The guests also share firsthand insights into recent investments, portfolio strategy, and the future of tech IPOs and innovation.

Key Discussion Points & Insights

Are We in an AI Bubble?

• Bubble Characteristics & Context
  • David Clark challenges the overuse of the term "bubble," urging for a more nuanced analysis based on historical precedents.

    "People are using the term bubble, you know, pretty indiscriminately. If you look back at historical bubbles, they have specific characteristics...it seems pretty clear to us that we are not seeing these things being applied into the AI market today." — David (00:04, 39:29)

  • The market is “frothy” with high valuations, but these are largely supported by extraordinary growth rates and real user adoption.
  • Distinction: Current AI boom is unique because the “installation phase” (infrastructure buildout) is occurring concurrently with the “deployment phase” (mass user adoption).

    "This is almost the first time we've seen a technology paradigm where the installation phase...is happening simultaneous with the deployment phase..." — David (00:52)

Case Study: Andreessen's Series A into Oboe (AI-driven learning)

• Investment Thesis & Product Fit (03:03–06:58)

  • Education is ripe for disruption: costs have continued to rise; AI can help people learn more efficiently.
  • What sets Oboe apart: vision ("helping a billion people learn a trillion things"), team pedigree (ex-Anchor/Spotify), product quality, and speed of execution.

    "You gotta look and feel and touch the product. And boy, when I use the product, it was smooth, it was quick and it was deep in terms of what I learned." — Bryan (03:57)

  • Competitive round and tight valuation process—good founders create their own investor markets.

    "We got to a point where we’re both equally, somewhat unhappy, if that makes sense." — Bryan (06:58)

• Due Diligence, Motivation & TAM Blindspots

  • Concerns included defensibility (could OpenAI or others build the same thing?) and the lack of $10B+ edtech precedents.

    "Is an open AI going to do that? ...The second piece is...is there any equivalent of a company like this becoming worth 10 billion?" — Jason (22:06)

  • Founders with relentless ambition and execution can "induce a market" even when TAM looks small or undefined.

    "Great founders build defensible businesses. Period." — Jason (22:06) "Maybe it's a white hole where you create..." — Bryan (23:55)

Setting & Understanding Valuations

• The Role of Breakneck Growth (07:19–08:41)
  • Early-stage AI valuations are high but justified by unprecedented growth velocity.

    "If you're getting companies that are growing to $100 million in 12, 18 months, that's an order of magnitude faster than we've ever seen before." — David (07:19)

  • Ownership targets for accelerators: around 7%, diluted over time, but big wins still produce substantial returns.

    "2% of a decacorn...those kind of returns can add up..." — Jason (08:48)

  • Power law dynamics are intensifying; funds must expect most gains from a handful of massive outliers.

    "...portfolio management is going to come into this...the power law is going to define returns even more than it has previously." — Jason (11:18)

Fund Returns, Portfolio Construction & Exits

• When Should VCs Exit? Balancing DPI vs. Home Runs (13:51–16:51)

  • Holding onto winners is crucial; premature selling of best companies can severely limit fund returns.

    “You don’t get to be one of the best firms globally by selling your best companies prematurely.” — David (14:21)

  • For liquidity management, VenCap prefers diversifying across stages and utilizing secondaries rather than pushing for early exits.
  • Timelines are now much longer—best companies may take 15 years from founding to full exit.

    “...for the best companies, at least sort of 15 year hold periods.” — David (21:13)

• Secondary Markets

  • LP secondary deal flow is way up post-2020 as institutions seek liquidity in longer-dated funds.

    "We've definitely seen an increase in secondary volume over the last sort of 12, 18 months." — David (18:55)

What Makes a Defensible & Scalable Startup?

• Market Creation > Market Size

  • Most transformative companies (Uber, Airbnb, Canva, 11 Labs) defied existing market size projections.

    "When Uber first came around...many asked, hey, what is the addressable market for black cars?" — Bryan (22:55)

  • Winning VC bets require believing an investment can “return the fund”—even if TAM is hard to quantify.

• How to Pitch Market Size?

  "A lot of times what we actually hope for is founder having sort of a world bending vision." — Bryan (25:46)

Startup Pivots & Founder Resilience

• Boom Supersonic: From Jets to Data Center Engines (30:48–39:01)

  • Hardware startups often “side hustle” into parallel verticals to keep the main business alive—a path followed by Tesla and now Boom.

    "Sometimes money shows up for a better idea...Sometimes you either need to do a pivot, you need to do some work for hire, you need to do some consulting to keep the lights on. I don't think there's any shame in this game..." — Jason (32:17)

  • The best founders adapt: "What we consistently see is the best founders find a way to make it happen." — David (34:16)

• Tech Giants as Power Companies?

  • AI/data center demand is so intense that major tech firms are investing directly in energy generation.

AI’s Margins, Business Models & Growth

• Gross Margin Evolution (50:45–54:17)

  • Early AI SaaS companies may have modest gross margins due to inference costs but are targeting high-margin segments and expect costs to drop over time.

    "If you have an extremely high margin, how should you think about that? Are you actually using a lot of the AI or not?" — Bryan (51:18)

  • Growth, not just margin %, will matter: gross margin dollars are becoming a better metric.

• Acquisition & Go-to-Market Innovation (57:25–61:51)

  • Traditional paid marketing channels have diminished effectiveness; founders must find creative ways—a global hackathon, alliances, influencer partnerships.

    "I am seeing a net new innovative classes of strategy that I have not seen before..." — Bryan (59:57)

• Momentum is the New Moat

  • In the fast-evolving AI landscape, the ability to build and ship quickly, AND distribute, trumps almost everything else.

    "The moat is always and always has been... velocity approach, product shipment as well as ability to distribute — these two must come together." — Bryan (60:21)

• Persistence Over Imitation

  "You'd have people make mistakes, you would have them learn, and it was great...what is always defensible is the relentlessness of the founder." — Jason (62:41) "Just, it's a marathon at a, you know, very brisk clip, and all you have to do is keep running. Most people can't run a marathon. Just understand that and keep running." — Jason (63:32)

Notable Quotes & Memorable Moments

On AI Bubble Fears

• “We are at the start of a major cycle here...as things stand to today, we are not in a bubble. That's not to say there won't be a correction. Corrections are part of a normal technology cycle.” — David (39:29)
• “Adoption is outpacing investment — demand is driving capital, not the other way around. This is the one I think is the most salient that you nailed...” — Jason (45:31)

On Founder Resilience

• “The best founders find a way to make it happen. And if that means pivoting or doing a side gig in order to get there, then that's what happens.” — David (34:16)
• “The relentlessness of the founder...their ability to inspire the team to ship products. ... Come back to me in year three.” — Jason (62:41)

On Market Size & Vision

• “We operate in an environment where it's sort of well understood that we may not be the best people to estimate the TAM...maybe we're not the best at estimating market sizes.” — Bryan (22:55)
• “What we actually hope for is founder having sort of a world bending vision...and a very good plan to do that.” — Bryan (25:46)

On Growth

• “If you're getting companies that are growing to $100 million in 12, 18 months, that's an order of magnitude faster than we've ever seen before.” — David (07:19)
• “Every investment, we take it very seriously as does this investment have the potential to return to fund.” — Bryan (24:41)

On Persistence and Competition

• “Momentum is the new moat — the ability to build product very quickly and actually know how to distribute, marrying these together, are the ones that are winning.” — Bryan (60:21)
• "For founders, just, it's a marathon at a very brisk clip, and all you have to do is keep running. Most people can't run a marathon." — Jason (63:32)

Timestamps for Key Segments

| Timestamp | Segment Description | |----------------|-------------------------------------------------------------------| | 00:04–01:01 | Are we in an AI bubble? | | 03:03–06:58 | Andreessen’s Series A into Oboe: rationale and process | | 07:19–08:41 | Valuation climate for AI startups | | 11:18–12:57 | Power law, portfolio construction, and exits discussion | | 13:51–16:51 | When and how to sell or exit fund positions | | 18:55–19:58 | Secondaries/Venture fund liquidity strategies | | 22:06–25:29 | Is market size/TAM a useful metric? What about world-bending vision? | | 30:48–36:01 | Boom Supersonic’s pivot to power/data centers and startup pivots | | 39:29–43:22 | Revisiting the question of an AI bubble and unique AI market growth| | 50:45–54:17 | Margins in AI companies and how founders should think about them | | 57:25–61:51 | Go-to-market, innovative customer acquisition, and moats | | 63:32–64:39 | Staying power and founder persistence as the ultimate competitive edge | | 65:31–67:57 | 2026 predictions: IPOs of “Mag 7,” AI-native products |

Closing Thoughts & Predictions (65:08–68:45)

• 2026 Hopes:
  • David: Hopes some of the “private Mag 7” (e.g. SpaceX, OpenAI, Anthropic, Databricks) will IPO, clarifying public vs. private market valuations. Puts an 80% chance on at least one going public.
  • Bryan: Excited for the emergence of truly AI-native products and UIs in the coming year (80%+ confidence).
  • Jason: Predicts a "golden age" of M&A is coming; expects several mega IPOs and wants to see young people pushed into entrepreneurship due to tightening labor markets.

Final Words

This episode offers a front-row seat to current venture thinking on AI, market cycles, founder quality, and the future of startups and innovation. The panel’s consensus: the best opportunities—and biggest wins—will go to startups led by relentless founders, building at record speed in new markets where demand outpaces even the most optimistic projections.