Shyam Sankar
speaker
319 appearances
2 recordings
2 series
first heard Apr 2025
last heard 6 May
Shyam Sankar’s voice in public audio — every appearance, attributed to the second.
Trend
recordings per month · last 12 monthsRecordings per month over the last 12 months — 1 in all, peaking in May 2026 with 1.
Appearances
In defense, we're starting to see that happen, where people are taking very sophisticated, the next generation of platforms, and they're starting with the software-first approach. What does this platform need to do? What are the threats? What do I need to be able to sense? What do I need to be able to affect? Can I even build that compute first? Which, by the way, goes pretty quickly.
You can measure that in months, not years, certainly not decades. Once I get that right, now I have the constraints around what I can build my airframe or my land vehicle or my submersible around because I know it can accomplish the essential mission that I need and has the capabilities to do it.
And you don't end up in these situations like with the frigate where you buy the frigate, you make a bunch of changes along the way, and you realize it won't float. The whole thing is for naught because you didn't start with the software-based approach. What about the boat? Yeah.
The same concepts apply to surface vessels and submersibles and to airframes, actually, where you're really thinking about, our big challenge in the Pacific is, how do we get inside the weapon engagement zone? The enemy has been able to really build up a lot of air defenses and hypersonics, and they can threaten and hold us at risk when we're inside of this engagement zone.
We don't have sufficient defenses. So maybe we need to build a bunch of relatively attributable, but certainly autonomous systems that don't have risk to force. that enable us to do the sensing we need to create the maneuver space we need.
And so there with Tenebris, it's really about using commoditized approaches to building ships so that we're not reliant on an industrial base that can't absorb more capacity and adding the mission-specific payloads that are highly exquisite and highly classified and modular at the point of need. And using that as, you know, it's not just about this ship
or this family of ships, but really a family of capabilities, whether it's space, air, sea, land, that give you this sort of software-defined approach to warfare.
I think software is a unique American strength, and it really comes down to culture. Look, there are smart people everywhere in the world. But think about it this way. There are zero Indian enterprise software companies that are competitive on the world stage. There are zero Chinese enterprise software companies. Europe only has a handful, and they're a pretty legacy. So what is it about...
America that we have all this software expertise. And it's cultural. Sometimes people look at Silicon Valley and they think, maybe we imported this culture from India or Israel. We didn't. We imported it from Iowa. It came from Bob Noyce, who is the co-founder of Intel, the co-inventor of the transistor.
And it comes from a Midwestern culture, a willingness to play positive sum games, open communication, transparency, a belief in working together. He brought that with him. The word open door policy, Noyce coined that. It's a Noycism. There's a way in which Silicon Valley doesn't even understand how much we are the progeny of this noise and this noise in culture.
And that's really permeated so much of how we organize ourselves to be able to build software at a scale that is world-beating by a long margin. The second best country in the world is Israel, and they can build canoes, not aircraft carriers. They can't get the culture of collaboration to scale even close to the level that we can.
This is why you see most Israeli founders, when they found their second company, they moved to New York to start it, because you're not going to get the big success that you want otherwise. And so then I think we should lean into that. You're not going to win the war by shooting bits at your adversary. At the end of the day, you have to bend metal.
Hardware is a critical capability, but we should do it asymmetrically. Our hardware should be way smarter than our adversaries. That's something that we have the capability uniquely to do. Even when you think about DeepSeq, which we have this Chinese AI model, you can see two things from it. They couldn't have built that model without essentially distilling or stealing this from OpenAI.
So you needed the requisite OpenAI investment. And then two, they did come up with a couple of very clever software optimizations, like novel innovative things that represent meaningful refinements on what's been happening here. But it is a refinement.
That as a precondition required, it's not a coincidence that all the frontier AI labs are within 50 miles, a 50 mile radius in San Francisco, right? That density of talent and the culture that's required to organize around it is an American superiority.
Thank you very much. Thank you very much. Thank you. Thank you. Thank you very much.
Okay. Thank you. Thank you very much.
Thank you. Thank you. Thank you very much. Thank you very much. Thank you. Thank you. Thank you.
Thank you. . . . . . ., en P. P. P. P. P. P. P. P. P,實,實, , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , a la in a
Showing 301–319 of 319 · page 16 of 16
← Previous