Autonomy unlocks the speed and scale needed to bend the economic cost curve of shipbuilding, since you can't out-compete China on material or labor rates so you must fundamentally redesign ships to use less steel and far fewer labor hours.
Dino Mavrookas argues that since the US can't beat China on cheaper steel or labor rates, the only path to competitive shipbuilding is redesigning ships from first principles for software autonomy and digitization, radically cutting steel and labor hours. ✦ AI generated
Dino Mavrookas · a16z Podcast · 2026-05-19 · original ↗
starts at this moment · 0:50
“How have you translated this alignment between Saronic's autonomy and the Navy's no man left behind ethos into contracts? And how do autonomy breakthroughs actually allow you to build more, faster products for the Navy?”
The real unlock to autonomy is speed and scale. Right? If I was sitting up here telling the secretary how we're going to build the same ships that we've been building for the last 50 years or the same ships that they're building in China, we're just going to do it better, faster and cheaper. I mean, he should just politely ask me to leave. You have to bend the economic cost curve.
verbatim transcript · starts at 0:50
0:50contracts? And and how does how do autonomy breakthroughs actually allow you to build more faster products for the Navy? >> So that's One, I'll I'll come back to what autonomy enables is and I'm a I'm a huge believer in this. We're unlocking capability that didn't exist before. Look, warfare will require humans involvement at some point but I've seen it and you should never send a human if you have the ability to
1:19send a robot. You just shouldn't do it. So let's leverage robotics and protect human life where we can. The real unlock to autonomy is speed and scale. Right? If I was sitting up here telling the secretary how we're going to build the same ships that we've been building for the last 50 years or the same ships that they're building in China, we're just going to do it better, faster and
1:46cheaper. I mean, he should just politely ask me to leave. You have to bend the economic cost curve. You have to build ships cheaper. You have to build ships faster. The only two levers you have to pull are material costs and labor costs. We're not going to buy steel cheaper than they're buying it in China. So, you have to take a first principles design to the ship and
2:10just fundamentally use less steel. How do you do that? You build for software autonomy and digitization. Same thing on labor hours. We're not going to acquire a per hour labor rate cheaper than they're paying in China. So, let's build ships much faster. For example, Marada's using is about a 50,000 labor hour on our first ship. Just for order of magnitude, put it in frame of reference. Apples and oranges comparison here. I'm
2:40going to acknowledge that up front. But a destroyer is 7 to 9 million >> Wow. >> labor hours. So, that's the scale and speed that autonomy unlocks because you can redesign the whole platform and make it much simpler to build. And then you build the Then you have the capability at the scale that you need to keep as many people as possible out of combat scenarios. >> You
- ·US cannot beat China on cheaper steel or labor rates
- ·Building the same ships faster/cheaper is a losing strategy
- ·Must fundamentally redesign ships to bend the cost curve
- ·Autonomy enables speed and scale in shipbuilding
- ·Redesign ships from first principles for digitization
- ·Software autonomy radically cuts steel and labor hours