Avi Bharadwaj

speaker

355 appearances 1 recordings 1 series first heard Jul 2026 last heard 7 Jul

Avi Bharadwaj’s voice in public audio — every appearance, attributed to the second.

Trend

recordings per month · last 12 months

Recordings per month over the last 12 months — 1 in all, peaking in Jul 2026 with 1.

Appearances

newest first · ▶ plays the moment

Talking AI · The VC's Lens: How AI Is Rewriting the Rules of Defensibility · 7 Jul 2026 podcast

approaches like state space machines or hybrids like Jamba and so on.

That could be better for use cases like edge devices because they are low memory or they are energy efficient or they can do real-time inference and so on.

So there is that niche where big model providers won't change their entire architecture to fix this use case.

So it's a fundamentally different approach.

And then the third bucket could be like unit economics.

Like there are lots of use cases where frontier models can do a good job, but the cost per token is so high that it may not make sense to use these big generalized models.

I was talking to a startup recently and they use a model called the Segment Anything model to pick out checkboxes from PDFs.

So apparently they have better accuracy and they have a lower cost.

As opposed to using a big frontier model, right?

So I think we see the world moving towards having dozens or tens of different foundation, like specialized foundation models for different use cases.

And like I said before, the trick is in building big platforms in which models are only one component.

Like you would use one or a handful of models, but there would be other layers in the stack that would make it pretty sophisticated.

So I think that's our view.

On generally specialized foundation models.

And your question on Figma, I think, is really interesting.

Thankfully, the private market companies don't gyrate like quite closely with public market stock.

Yeah.

Otherwise, it would have been crazy to to manage our investment thesis.

So we need to have our investment thesis needs to be looking out five, 10 years out and thinking about what will happen in the medium to long term.

But there are trends, I think, that that way.

Showing 161–180 of 355 · page 9 of 18