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
I think it's one example of it. It'll cut out. It'll make you a much smaller, leaner fighting force for sure. So let's just call it half. Yeah. You can probably make everything half as big and as lethal or more lethal than before. There's so many things with tanks. How do you clear the range of fire? The commander has to pop their head up.
They have to somehow get on a phone and tell the other tanks. When the soldiers are dismounting, you don't want to be in the field of fire. That should all be software. No one should be picking up the phone. The tank should be able to communicate to the other tanks.
So once you start unpacking the layers of how many things we're just covering down with human labor, you start to realize, oh, this stuff's going to go away. The humans are doing it slow. We're doing it at the speed that humans can do it. And we make mistakes when we're doing it. And it has a huge training burden to do it. So how long does it take to even ramp people up changes?
So I think it's going to be a pretty dramatic impact. You have to earn it, though. You have to kind of ship these innovations to the operating fighting force. And no one will be faster to adopt this than the fighting force. It makes them more survivable, more lethal. There was one project we did. At NTC, at Fort Irwin, the division was roaming around the desert.
You know, division, big army formation. But using the latest technology, just like this, they were able to get their division footprint down to 26 people. So instead of the normal 400-plus people you need, a big jock, big screens, which means the enemy can see you, you're pretty immobile, you're not going to be survivable, you're going to be blown to smithereens.
26 people, the red team could not find them in the desert.
Well, you definitely want first contact to be made by... entities that have no blood in their system, right? Like the front line of troops, but there's also the front line of robots, and that should be further front. And we need to be thinking about how we have less risk to force as a consequence and have, you know, improved mission outcomes by doing this.
I do think, though, that the tail is where the immense opportunity is. So the tail here, so yes, it's going to make the tooth sharper, and you're going to need fewer teeth. But the tail... is enormous. You have more people in the Acquisition Corps than you do in the Marine Corps. And the human processes that we use to support all of this is both very latent and slow.
It obfuscates indirectly just because humans are touching and massaging the data, like what's actually going on. And we need to integrate that. When I think about the most important missing strategic asset to the US government from a deterrence perspective, it is an integrated view from the factory floor to the foxhole.
And by that, I really mean, let's articulate the threats we care about under the timelines. These are the foxhole scenarios that matter to us. This is the national defense strategy. This is who we're trying to deter and when. And work that backwards to, based on what I'm spending and building, does any of that stuff matter? Will it be there in time?
And I think the Department of Defense is the only institution I know of that has divided up supply and demand. The process of integrating supply and demand is usually the beating heart of a commercial company. This sales and operation planning process, people have different names for it. In the department, we ask the services to go build things. That's the supply side.
And we ask the combatant commanders to do the war fighting. That's the demand side. They have to respond to real world events. They don't really get that much of a vote in terms of what's happening here. We have a lot of control over what's happening here. We have ideas about the scenarios in the world we care about making sure don't happen.
How do we integrate these things so that we understand like, you know what, it actually would be better to buy a less capable weapon system we can make seven times faster at a lower price point because the other stuff's not going to be ready in a time that's relevant to anyone, right? It's just running headless by doing this.
Now, and I think if you had that, it becomes the first way you could have meaningful conversations like, yes, I know there are jobs in your district. But do you understand how you're actually going to sacrifice American prosperity? We are going to lose over 100 jobs in your district. And I think that becomes now you've changed. Turn the tables on what is politically survivable.
Yeah, the Titan truck is a satellite ground station on wheels that's really about enabling long-range precision fires. So how can we do deep sensing of the battle space to enable precision fires of the enemy deep in their territory? What I think is really cool about the Titan program is that we were the first software company to win as a prime contractor. It is a physical thing.
We still don't think of ourselves as a hardware company. We built a team of super friends, of power horses, Andral, Northrop Grumman, L3 Harris, Sierra Nevada. They're helping us build this thing. But what makes the truck in some sense is a commodity.
What's in the shelter is what's doing the AI that's finding the enemy on the battle space using national technical means, using satellite communications, doing the processing, processing, exploitation, and dissemination on board of this with a two-man crew. So very, very efficient. and enabling then the transmission to the gun line to actually go destroy and close with the enemy.
And building this around the software first. What does the software need to do? That is our lethal capability. Then how do I make it as, amazing for the warfighter as possible as a consequence. Rather than doing what we've done backwards so many times is we have this hardware platform, and then we're constrained by what software can we put in there.
And it's not good enough, and it's not adaptable, and there's not enough compute. You kind of designed it backwards. And this is one of the paradoxes. So if you're a true historical hardware company, you can think about like building a house. If you're building a house, you've got to build the foundation first. You are entirely limited by the foundation. Software is not like that.
With software, you've got to build the kitchen first, and you build the rest of the house around the main thing. And now we're starting to see that culture enter into defense. That's what Tesla does. That's what SpaceX does.
Showing 281–300 of 319 · page 15 of 16
← Previous
Next →