The AI wave will produce enormous roadkill — 'it's like Hollywood, man, 95% are not going to be there' — but seed investing still works because you don't need to be in the one huge winner if you own 5% of a $2.6 billion outcome, so seed is not dead, just crowded and commoditized.
David predicts heavy casualties from the AI boom (analogous to Hollywood's 95% miss rate) while defending seed's durability: a 5% stake in a median $2.6B company returns a fund, which is why seed is crowded and commoditized rather than dead. ✦ AI generated
David Frankel · 20VC · 2026-08-08 · original ↗
starts at this moment · 20:17
“Is this like peak bubble? ... Are we headed for another crash?”
will there be roadkill from this wave? Oh my god, there's going to be a lot... How many times, Harry, over the last 11 years have you heard this is different? This is different... It doesn't mean that there aren't survivors and companies that are going to change the trajectory of technology forever... these are the metas and the Googles of our era. Highly likely. But wow, like it's Hollywood, man. Like 95% are not going to be there. And it goes back to why is seed interesting? Like I don't have to be in the one... if you have 5% of a $2.6 billion outcome, you've returned your fund... And that's why I think seed isn't dead. I think seed is crowded and to some degree very commoditized.
verbatim transcript · starts at 20:17
20:17mean that there aren't survivors and companies that are going to change the trajectory of technology forever. And I think in, you know, open AI and anthropic and SpaceX, we're seeing that already. Like these are the metas and the Googles of our era. Highly likely. But wow, like it's Hollywood, man. [laughter] Like 95% are not going to be there. And it goes back to why is seed interesting? Like I don't have to be in
20:44the one, right? Like if there were five companies so of that that are worth $5 trillion literally you go SpaceX I mean I'm not even including I'm saying like um with exits so if you look at SpaceX Tesla Meta that's trillions of dollars already you take then Nvidia I think Nvidia started pre25 years ago but even if you look at the last 25 years ago you can add Palanteer to that PaloAlto
21:12networks that's about $5 trillion of market cap And then the the other 495, right, at a 2.6 billion average. And some of those are, you know, we hope everything looks like Shield AIO. But if you have 5% of a 500 million, if you have a 5% of a $2.6 billion outcome, you've returned your fund. If you have a $500 million outcome, it's incredible still. And that's why I think seed isn't
21:40dead. I think seed is crowded and to some degree very commoditized. I feel commoditized. I've said this many times. I feel like brand and in some in some regard distribution as in your portfolio and people saying nice things about you get you to the table. >> But if it's commoditized, does price not just become the separator? And if price is the separator, the mega platforms win. >> Well, the problem is the mega platforms
22:07are taking call options. So, is this good for the mega platforms? Is this good for the LPs? Or is this good for the entrepreneurs? Well, probably for 95% of entrepreneurs, it's not good. Why? You get more I'm I agree with you, but I'm just playing devil's advocate. You get more money at a higher price with mostly a more junior VC who will let you do your work and not get in the
- ·The AI wave will produce enormous roadkill
- ·Hollywood analogy: 95% are not going to be there
- ·Survivors will be the metas and Googles of this era
- ·Seed investing works without betting on the single winner
- ·A 5% stake in a $2.6B outcome returns your fund
- ·You don't need to be in the one huge winner
- ·Seed is crowded and commoditized, not dead