I just wanted to get your two cents on where we are in terms of, uh, gross margins in AI startups.

Look, I think margin, uh, you know, I, it's, it's funny.

I used to run growth and used to be a CFO, so margin is a near and dear, uh, topic to me.

Um, there are a couple ways to think about it, where one is that we all know that AI comes with inference cost, so the more you use it, there are some marginal costs baked into it.

So when you think about a true AI native product, if you have an extremely high margin, how should you think about that? Are you actually using a lot of the AI or not? So that's, like, one way to think about that question.

The other way to think about this question is, look, what we see as a margin at the end of the day is aggregation and averages, like on average when you sell a product, what do you take away? A- and what actually the truth is what we're seeing is, it depends on different slice of cus- customers.

So when you go deeper into the product itself and u- usage pattern itself, you have prosumer users, you have enterprise users, you have creator class.

Um, and so, you know, whatever way it takes to acquire customers and- and- and to- to build that market dominance, if you're a founder, that's something you really need to think through.

Talking about Eleven Labs, how did you first connect with them and how did you get your conviction?

Another really, really fun story.

I have a portfolio, founder Gaurav from Captions, actually my first investment ever.

Okay.

And I trust him.

He has a very strong technical acumen.

He's an extremely strong product thinker.

How do you think about that? Are you seeing that as a much more attractive way to invest into companies, look for those companies that have really clean infrastructure?

Brian, why don't you share your story of how you got excited about Clulee of how you and Roy met and build this relationship and how this partnership formed?

I had a good contact in New York who travels in the young crack folks area.

And her name is Ali Deboe.

And one of the lists that she sent had Roy in it.

And I like read what they're working on.

And it sort of reminded me of, oh, like a scout thing or it's on the edge of virality.

I'm like, oh, this is interesting.

Yeah,

Mm-hmm.

And sort of think o- thinking about that, I think there is a beginning of a lot of the consumer products and products that are used by many people have a funny beginning sometimes and I think...

So one calculated risk side, we're comfortable with that.

And I think second thing is when I...

You know, I have a saying where momentum is a moat today in this sort of era of AI applications.

And when I say momentum is a moat, I don't mean just distribution.

I also mean product iteration.

Mm-hmm.

So I think the optimistic view is that the models are now mature enough and many are available either open source or via API for people to build great, more traditional consumer products on top of them.

it's interesting that you asked that question because I was thinking about the past, what, 15 years, 20 years, where, as you said, like Google, Facebook, Uber, all the names.

Yeah.

And it's interesting because when you think about internet, mobile, cloud, everything together, there were all these amazing names.

And you ask, has that slowed down? I think the cloud, mobile, all that had a lot of maturity baked in.

Like the platform was around for like 10, 15 years.

Every little nooks and crannies has been explored to some extent.

What's going to get us to something new here in the realm of connection? Or has that just been one?