Precursor Ventures wonders why AI startups are raising bigger rounds than ever when AI should make them cheaper to build. Past startups like Uber and Lyft burned cash on customer acquisition. Today's AI companies are raising more money than their compute costs justify, and it's unclear where all that capital actually goes.
Chapters
And so I think in Uber, Lyft And Instacart and DoorDash, those hundreds of millions of dollars were going into customer acquisition, and we're building user moats and network effects that really enhance the businesses. I continue to wrestle with this paradox, which is that, oh, AI, in theory, makes these companies much cheaper to start. I'm not sure it makes them cheaper to build, but we'll go with the premise. It makes them cheaper to start, and it makes them more efficient. Then I'm like, well, then why are they raising so much money, and what are they going to do with it? And to me, it's unclear. Yes, you need to fund the purchase of credits, or let's call it compute. You have to finance that. But of these companies, it feels like the money they're raising is growing faster than compute costs. So.