Give costly AI features away free to grow fast

David Barnard… have the potential to be a multi-billion dollar outcome or does it... You can correct me if you're thinking about it differently, but then how do you think about the investability of these consumer apps? Because we are seeing apps get to 10 million super quick, 20 million super quick, but are they going to go from 20 million to 200 million to 600 million to be these next Duolingos or Life 360 or these other big publicly traded subscription apps?

Olivia MooreThis is something we think a lot about and I think there's a couple layers to it. One would be this concept of what I call giving away dollars for pennies, which would be that a lot of startups now are achieving really fast growth by essentially subsidizing the model costs, the end consumer. So giving away Nano Banana Pro for free, things like that. And you can get very fast growth that way. To me, it's almost similar to the on-demand economy like Uber, DoorDash, a trillion other apps, most of those the unit economics catch up to you eventually. In some cases you're able to achieve escape velocity almost and then make users profitable. But I think you have to be quite, quite good at marketing and distribution and really lock users into the product if you're going to grow that way. The other element that we think about a lot is what the big labs and model companies are doing and what the incumbents are doing.

Olivia MooreAnd how does that impact you as a standalone consumer product. We're seeing Claude go extremely hard on Prosumer, ChatGPT launching their own health product, trying to own more of their own app store. Gemini is doing all the personal intelligence stuff, and so we think a lot about where those companies have the right to win versus where an independent startup can build something much more opinionated and almost verticalized, and retain and …