Ben Horowitz

You're going to run into the wall with the underlying technology.

And you're going to have to keep moving.

I think defensibility just in general is very tricky when...

For example, like a model for a humanoid robot is like a pretty hard technical problem still, despite everything.

So like, that's one thing.

Look, you know, possession of the customer is still very, very valuable.

And like an easy example of that is ChatGPT.

Like how much model differentiation does OpenAI have at this point?

You know, maybe they have negative differentiation in some areas,

But they've got the most consumers and that still matters and that kind of can get them to the next square.

Yeah, and the brand.

So those things, I think as long as the

customers are humans.

That's still quite a big thing.

I really think that knowing like, being able to articulate

really why you're building this company in a way that convinces even you that you should rejoin your company is really a thing.

So you want to, if you can convince yourself like completely, like this is the greatest fricking idea in the world, then that's the most compelling thing.

I think that the mistake people make in fundraising is they go, well, what does Ben want to hear?

And does he want to see a graph like this or does he want to,

you know, know this about market size or whatever, that's very hard to figure out and you usually get it wrong.