Why the Pentagon Wants 240X More

It’s the Battle of Agincourt on October 25, 1415.

Henry V of England wasn’t one to watch his men from a hilltop.

He swung his sword in the mud alongside them.

Overhead, thousands of English arrows rained down on France's heavily armored knights.

Cheap weapons fired by commoners crushed some of the most expensive soldiers in Europe.

During the fighting, an axe struck Henry's helmet and damaged his crown.

Back then, leaders had skin in the game.

But look up.

The future was whizzing overhead.

The Distance of War Widens

Over the centuries, generals moved behind the lines. Then off to headquarters far beyond the battlefields.

Pilots flew at 30,000 feet. Drone operators now work in trailers thousands of miles away.

Today, computer-vision models can recognize targets, track their movements, and guide autonomous drones toward them.

Software on one side. A human on the other.

There's one more potential human in the kill chain, too.

You. The potential shareholder, investing in the companies behind this technology using an app on your phone.

So it goes.

My grandfather was a Marine who fought in the Pacific during World War II. He knew what war cost. He made sure I did, too.

But as a tech investor, I need to understand what's changing, and who's being paid to build it.

This is the first of four issues about how computer vision and AI will change warfare.

Today, we'll start in the air.

After that, we'll cover China’s machine-gun robot wolves…

UAPs emerging from the sea…

And finally humanoid robot super soldiers.

Agincourt, Again

America is protected, in part, by two enormous oceans.

Poland, Taiwan, and Israel don't have that luxury.

Oceans don't stop missiles or drones, but they make an invasion of the United States extraordinarily difficult.

In January, the Pentagon released its 2026 National Defense Strategy.

The new U.S. strategy prioritizes defending America and the Western Hemisphere as well as deterring China in the Pacific...

And requiring old allies to take on more of their own defense.

It also includes rebuilding U.S. weapons production.

It's not a simple withdrawal from the world. It's a change in who carries the burden.

Consider the flashpoints: Russia and Ukraine in Eastern Europe. China and Taiwan in the Pacific. Iran, Israel, and their neighbors in the Middle East.

What this means is European, Asian, and Middle Eastern allies may buy more autonomous equipment, right alongside the Pentagon.

On September 30, the War Department announced a new four-star Autonomous Warfare Command, intended to ramp up the use of drones, robotics, and AI across the armed forces.

Its planned launch is October 2027.

The interim project has a familiar name…

Project Agincourt

The Pentagon's autonomous-warfare office received roughly $226 million in fiscal 2026.

Its fiscal 2027 request jumped to $54.6 billion. That's 240X!

Only about $1 billion is in the regular budget. The other $53.6 billion depends on additional legislation.

But given the evolution of technology, it's not a question of if this much money is allocated...

It's only a question of when.

So, now is the time for investors to build a watchlist of companies that could benefit when that money floods in…

Because some of these stocks are aiming to become the Lockheed Martins, General Dynamics and Northrop Grummans of tomorrow.

Asymmetric Drone Warfare Is Here

On June 1, 2025, Ukraine carried out Operation Spiderweb.

Its intelligence services concealed small drones inside wooden cabins loaded onto trucks and driven deep into Russia.

Then they launched them toward military airfields.

Ukraine said 117 drones participated. Some targets were more than 4,000 kilometers from the front.

Think about the math on a single hit. A $1,000 drone against a $100 million bomber.

That's a 100,000:1 cost-exchange ratio.

Today, both sides use drones for reconnaissance, attack, and strikes on infrastructure.

Electronic jamming has also pushed engineers to give more drones the ability to navigate and track objects autonomously.

That doesn't mean every drone chooses its own targets or fires without a human...

But the direction of the technology emerging from this is clear.

Machines are learning to see and making their own decisions...

Much like a Tesla does driving down the road.

How a Machine Sees

If you've ridden in a Tesla, you've seen how computer vision works on its display: cars, pedestrians, lane markings, traffic lights.

The system starts with cameras taking images over time.

A neural network identifies objects and estimates where they are.

Another part of the software tracks them over time and classifies them.

A detection might say: Person. Confidence: 94%.

In a car, that information may contribute to a braking decision.

On a military drone, the consequences are lethal.

Machine sensors also aren't limited to human eyesight.

Thermal cameras detect heat. LiDAR measures distance. Radar can detect objects in conditions where visible-light cameras struggle.

Sensors can monitor 360 degrees without getting tired.

But they can also fail, be fooled, or misclassify what they see. Fast detection isn't the same thing as sound judgment.

That's why this is an entire industry, not just a camera attached to a drone.

Someone builds the eyes. Someone builds the processors. Someone writes the software.

And someone pays for them.

According to Fortune Business Insights, the global military-drone market is estimated at $22.49 billion in 2026...

And projected to reach $52.31 billion by 2034.

That's an 11.1% annual growth rate.

Three Flying Drone Stocks I'm Watching

AeroVironment (AVAV): An Established Supplier

AeroVironment makes the Switchblade family of loitering munitions, along with reconnaissance drones and other defense systems.

Its newer Switchblade technology includes automated target-recognition capabilities, with human oversight still an important part of its military use.

In its latest quarter, the company reported $480.5 million in revenue, up 6%, and $1.5 billion in funded backlog orders.

Management expects $2.125 billion to $2.225 billion in fiscal 2027 revenue.

Of these three stocks, it has the best financials.

The question is whether larger U.S. purchases and orders from our allies translate into sustained production growth.

Risks include contracts arriving late, manufacturing bottlenecks, and the complexity of acquisitions.

Red Cat (RCAT): The Small-Drone Manufacturer

Red Cat makes the Black Widow, a small surveillance drone chosen for the U.S. Army's Short Range Reconnaissance program.

The company is also gaining overseas customers.

It announced orders involving NATO and Asia-Pacific allies, including Japan, in 2026.

Second-quarter revenue reached $20.2 million, up 527% from the year-earlier quarter.

That sounds spectacular. And it's genuine growth from a tiny base.

Its gross margin was still only 16.1%, so scaling production profitably will be a future challenge.

Red Cat also works with outside AI partners, so its value isn't just in a drone that flies…

It's in a platform that carries increasingly capable perception software.

Safe Pro Group (SPAI): The Software Moonshot

This is the smallest and riskiest of the three.

Safe Pro uses computer vision to examine drone imagery and identify landmines and unexploded ordnance.

Same war. Same cameras. Same object detection.

But instead of finding targets to attack, its software helps find explosives so people can avoid them.

In August, Safe Pro reported more than $5 million in 2026 AI-related orders.

On October 7, it estimated more than $2.5 million in third-quarter revenue.

Those numbers are tiny. And orders aren't the same as recognized sales.

But a June order supporting the U.S. Army combined Safe Pro's AI analysis system with Red Cat's Black Widow drones.

That gives us something worth following: software sold alongside real military hardware.

The question is whether small contracts grow into repeat licenses and wider deployments.

Waiting for the Money

Three stocks. Three different bets.

AeroVironment (AVAV) has scale and funded orders.

Red Cat (RCAT) has a fast-growing small-drone business and new potential customers.

Safe Pro (SPAI) is trying to turn specialized vision software into recurring defense revenue.

Technology has direction. Step back far enough, and you can predict it.

Farther. Cheaper. More precise. Fewer humans are exposed to the most dangerous tasks.

Henry V swung a sword at Agincourt. The man facing him could swing back.

Six centuries later, the soldier on the ground still has skin in the game.

But the machine watching him from above has none.

Just carbon fiber. Circuits. Silicon. And code.

So it goes.

Next issue: China's robot wolves and other fighting machines on the ground…

Always be prospering,

Harry Seldon
Moonshot Minute

P.S. BREAKING NEWS: Yesterday, Ukraine announced a new milestone in air drone warfare. One of its drones flew a wounded soldier out of a front-line position cut off by mined ground and Russian drones. Ukraine's military calls it the first rescue of its kind. And it was all captured on video.

🔓 Premium Content Begins Here 🔒