Transcription
Everyone knows there are 6 branches of the U.S. Military. The Army. The Navy. The Marines. The Air Force. The Coast Guard. The Space Force.
And everyone assumes American power gets projected one way. Through multi-billion dollar aircraft carriers. Stealth fighters. Marine battalions kicking down doors.
But take a close look at the Pentagon's most recent budget request. There’s something… not quite right. An anomaly. A massive pile of cash, larger than the entire annual budget of the U.S. Marine Corps. It doesn't belong to the Army. It doesn't belong to the Navy. And it doesn't belong to the Air Force. It funds an office most Americans have never heard of. The mission of that office is to fight the next war over Taiwan. And the replacement is already being built. It's called the Defense Autonomous Warfare Group, or DAWG for short. It was built in secret. Was never voted on. And is the future of US warfare. The question isn’t whether it works… it’s who it answers to when it does.
When the Pentagon released its nearly $1 trillion budget request for fiscal 2027 - an amount that’s more than the entire GDP of Switzerland - defense reporters did what they always do. They flipped to the part about carriers and fighter jets. They scanned the totals and went home. But buried inside that document was a line item nobody expected: The F-35 program is being trimmed in key areas. Carrier construction is slipping further to the right. Older Reaper drone squadrons are being divested faster than first planned. Legacy systems are being delayed, cut, or restructured. But somehow, the total budget went up. That kind of growth doesn't happen in defense. Not in peacetime. And definitely not at this scale. You can’t trim fighters, delay carriers, slow shipbuilding and retire older systems… while increasing spending to historic highs. The arithmetic only works one way. A massive amount of cash is being shifted out of the legacy arsenal and funnelled into something else. Something that doesn't appear in any of the line items. Defense industry trade publications missed it. The cable news segments missed it. Even the congressional staffers who wrote the markups missed it. Because the money wasn't sitting where they expected it to be. So if the Pentagon is cutting pieces of the old force while spending more money than ever before, an obvious question follows. What are they actually building?
What made the anomaly so strange wasn’t just the size. It was the speed. Major Pentagon budget shifts usually happen slowly, over years and sometimes decades. Entire congressional careers are built around protecting existing programs from disruption. The F-22 took years to scale. The Ford-class took years to justify. Even cancelling old systems normally becomes a political knife fight. But this money moved differently. Almost as if the Pentagon believed it was running out of time. That changes the meaning of the budget entirely. Because bureaucracies only abandon inertia under two conditions: catastrophic failure… or imminent threat.
Every dollar in the Pentagon belongs somewhere. Every department, from the Army to the Space Force, and the Coast Guard get their slice. But Special Operations Command, or SOCOM, is different. It runs the country's most sensitive missions. Counter-terrorism. Hostage rescue. Direct action against high-value targets. It has its own acquisition authorities, its own procurement pathways. There’s a long history of fielding systems nobody outside the building hears about until years later. Buried inside that command, there was an office that had a budget of $225.9 million in 2025. For fiscal year 2027, the Pentagon has requested $54.6 billion for that same office. That number isn’t a typo. A 24,000% increase. In a single budget cycle. Twelve months. It's the largest single year-over-year boost of any program or office in the entire Pentagon. Compare that to the Marines. Their request for the same year landed at $52.8 billion. Every Marine on every base around the world. The helicopters. The amphibious ships. The salaries. The training. All funded at roughly the same scale as the anomaly.
The same budget request contained another first: a dedicated breakout for autonomous systems across the entire Department of Defense. Billions for aerial drones, autonomous surface vessels, ground robots and the software tying them together. A pile of cash like that doesn't appear by accident. Everything about it is intentional… the destination, the location, the amount. Someone in the Pentagon is signing the checks… and they’ve decided the next war can’t be fought with the inventory. It’s just not enough.
Follow the money and it leads to a program called Replicator. Most Pentagon programs have names designed to be forgotten. Acronyms stacked on acronyms. But Replicator is different. The name is a tell. It means exactly what it sounds like. A system designed to produce thousands of identical units at scale and at speed. Replicator was announced in August 2023. The goal is to deploy thousands of attritable autonomous systems across multiple domains within roughly 2 years. "Attritable" is the key word. In Pentagon speak, it means something the military expects to die. A weapon cheap enough to lose. The opposite of an F-35.
Traditional Pentagon procurement runs through JCIDS, the Joint Capabilities Integration and Development System. This is where the validation panels and testing reviews happen. And the countless congressional oversight committees. The normal timeline for a major U.S. weapon system is 7 to 10 years. The F-35 program was launched in 2001. The first squadron didn't reach initial combat readiness until 2015. 14 years. Replicator targeted 18 to 24 months. To hit that timeline, Replicator skipped JCIDS altogether. It used special acquisition authorities the Pentagon usually reserves for emergencies. The Pentagon promised thousands of systems by 2025. Analysts tracking the program concluded only hundreds materialized at meaningful readiness levels. This was down to hardware problems, supply chain issues and software failures among other things.
Then Ukraine changed everything. Cheap quadcopters began destroying tanks worth millions. What began in 2022 as an expected Russian conventional victory instead turned into a drone-led war: sea drones pushed back the Black Sea Fleet, and decentralized production outpaced traditional defense contractors. By 2025, Ukraine was reportedly producing up to 100,000 cheap attritable systems per month. A drone that costs less than a used car was destroying weapon systems built over decades. That realization shattered decades of Pentagon assumptions. Because the same equation could work in reverse. Against American carriers, jets and tanks.
So, instead of killing Replicator after it missed targets, the Pentagon expanded the idea. Replicator was not the final program. It was the prototype. And prototypes exist for one reason. To test whether an entirely new model can survive contact with reality. The Pentagon learned something during Replicator that mattered more than the missed deadlines. It learned Silicon Valley could iterate weapons faster than the traditional defense sector could approve PowerPoint slides. A startup could redesign hardware in weeks or push software updates overnight. Traditional procurement was built around static systems that stayed operational for decades. Autonomous warfare changes too quickly for that model to survive.
Autonomy changes what a weapon actually is. For most of modern military history, the platform was the centerpiece. The tank. The destroyer. The stealth bomber. The physical machine carried the value. Now the software matters more than the chassis carrying it. A drone frame can be replaced in hours. The autonomy stack coordinating thousands of systems cannot. Which means the Pentagon is beginning to treat code the same way it once treated nuclear submarines: as a strategic advantage capable of determining the balance of power itself.
Anduril. Shield AI. Saronic. Skydio. Most were founded by software engineers and startup founders, not traditional defense executives. Anduril is the clearest case. Founded in 2017 by Palmer Luckey after selling Oculus to Facebook, the company moved from virtual reality into defense technology almost immediately. Within years, it was winning contracts measured in the hundreds of millions. Then billions. In March 2026, the Army awarded Anduril an enterprise agreement worth up to $20 billion over 10 years. The company isn't centered around carriers or tanks. Its core product is a piece of software called Lattice. It’s designed to fuse sensor feeds from drones, radars, satellites, cameras, and autonomous platforms into a single operational picture. A coordination layer for swarms. Saronic is building autonomous ships. Shield AI is developing AI pilots. Skydio scaled tactical drone production after Chinese-made systems were pushed out of operational use. Together, these companies formed something the Pentagon had not possessed before. A parallel defense industrial base optimized for autonomous warfare instead of Cold War procurement cycles.
Venture capital quickly followed. Billions of dollars flowed into defense tech between 2022 and 2025. Funds that once chased consumer apps began chasing autonomous warfare startups instead. The big primes are not gone. They will not disappear. But the procurement gravity has shifted. The Pentagon's most urgent dollar is no longer flowing to Bethesda or Arlington. It is flowing to Costa Mesa. Austin. San Diego. And a 5-million-square-foot facility in rural Pickaway County, Ohio called Arsenal-1. That is where Anduril began producing Fury combat drones in March 2026. 3 months ahead of schedule. The industrial center of gravity began moving away from giant shipyards and toward robotics factories, software campuses, and drone assembly lines. And the Pentagon's personnel system began changing too. It began rotating software engineers through a single defense assignment platform called GigEagle. Not soldiers. Coders. Roughly 4,500 engineers were on rotation at launch in April 2024, rising to over 13,000 by mid-2025. The Pentagon was no longer just recruiting soldiers. It was beginning to recruit bandwidth. Not mass mobilization. Network mobilization.
But what is all this for? What kind of enemy demands a force that looks more like a startup than an army? The answer is in the Pacific. For decades, American planners assumed the Pacific belonged to the United States Navy. The Seventh Fleet has been the dominant force in those waters since 1945. That assumption died sometime around 2018. China now fields the largest navy on Earth by hull count, more than 370 ships. The U.S. Navy has less than 300. The gap widens every year. Chinese shipyards can build a destroyer in roughly half the time American shipyards can.
But ship count is not the real problem. The missile is the problem. China spent years developing anti-ship ballistic missiles, the DF-21D and DF-26. They are designed to do one thing. Hit an aircraft carrier in open ocean from over 1,000 miles (1,609 km) away. If those missiles work as planned, every American carrier operating close to Taiwan becomes a target almost immediately in a major war. A Ford-class carrier carries thousands of sailors. Losing one would be catastrophic. This is why the Pentagon stopped trying to match China ship for ship. Geography makes the trap unavoidable. Taiwan sits roughly 100 miles off the coast of mainland China. Any invasion fleet crossing the strait creates a short vulnerability window before troops establish a beachhead. But traditional American power projection struggles inside the missile envelope. The closer American forces move toward Taiwan, the more exposed they become. The farther away they remain, the harder it becomes to intervene in time. The old model breaks under the geography.
The only thing that closes the gap is mass. The United States is no longer preparing to win the Pacific with carriers alone. It is preparing to flood the battlespace with machines. In June 2024, Admiral Samuel Paparo described the strategy publicly in a single phrase. Hellscape. "I want to turn the Taiwan Strait into an unmanned hellscape.” If China launches an invasion, the United States floods the strait with autonomous systems immediately. Thousands of them… just waiting. The goal is to make the Chinese decision to invade so painful that Beijing changes its mind. They want to buy the Taiwanese military enough time to prepare and dig in. They want American conventional forces to move into position without external pressure.
The numbers required to make Hellscape work are staggering. The strait holds enough autonomous systems that no Chinese vessel can move 300 feet (91 meters) without crossing a sensor or a warhead. Imagine the surface area of Manhattan, every park, every street, every rooftop. Now cover every square inch of that island in a layer of small autonomous machines that refreshes every 10 minutes. That is the density Hellscape demands. Autonomous warfare on that scale has no precedent. Ukraine demonstrated fragments of this future. Hellscape scales it into an industrial system. And industrial systems create a new kind of military requirement. Sustainment at machine speed. A carrier strike group can deploy for months before needing maintenance cycles. Autonomous warfare burns through hardware continuously. Drones crash. Sensors fail. Batteries degrade. Electronic warfare corrupts software and burns out components. Hellscape only works if replacement systems are constantly flowing into the theater faster than the enemy can destroy them. Factories become part of the kill chain. Supply chains become strategic infrastructure. Software engineers become operational assets.
Suddenly the budget numbers make sense. The Pentagon is no longer funding prototypes, it’s funding the next generation of warfare. It’s funding production. Because autonomous warfare only works if the systems already exist before the war begins. You cannot produce swarms after missiles start flying.
But this is where the plan runs into a wall. Every American military structure was designed around human decision speed. A Navy carrier strike group is run by a single Admiral. An Air Force fighter wing is run by a Colonel. Fighter squadrons. An Army battalion is run by a Lieutenant Colonel. They were all built around chains of command measured in minutes and hours. Hellscape does not operate on human time. If thousands of autonomous systems are moving simultaneously across the Taiwan Strait, no admiral can individually authorize every engagement. The systems must identify threats, coordinate responses, and act within milliseconds. That requires an entirely different command architecture.
In a traditional military structure, doctrine changes slowly. Orders move downward through ranks. Autonomous systems can change behavior through software updates pushed globally in minutes. A targeting model rewritten in California can alter battlefield behavior in the Pacific before sunrise. That means future wars may evolve faster than humans can process them. Not because the machines became conscious, but because software cycles are now faster than military bureaucracy itself. The Pentagon's own internal doctrine on autonomous weapons requires what officials call "appropriate levels of human judgment over the use of force." That phrase was deliberately written to be vague. The translation is straightforward. A human Admiral can sign off on a set of rules. Engage any vessel within this geographic box that meets these classification criteria. The drones then make individual engagement decisions inside that box. That is the only way to make swarm warfare work at speed. It is also a hard break from how the United States has authorized lethal force for the entirety of its modern history.
But there is a legal problem. Creating a new branch of the U.S. armed forces requires an Act of Congress. Not an executive order. Not a Pentagon directive. A bill. Passed by both chambers. Signed by the President. The most recent example is the Space Force in December 2019. No equivalent legislation exists for autonomous warfare forces. No public hearings established a seventh branch. No floor votes. No formal service authorization. And yet the Pentagon has assembled the budget. It has hired the personnel of a tech company. It has written the war plan for the Pacific. Which raises the real question. If Congress never authorized a new branch… what exactly is this thing?
The answer hides within the Department of Defense organizational charts. The Pentagon has a workaround. It's called a Joint Task Force - a command structure that doesn't need Congress to exist. It’s the same mechanism that built Joint Special Operations Command in 1980. That command still exists today and the Pentagon is using it again. And DAWG is exactly that workaround. The $54.6 billion office sitting inside Special Operations Command. The hidden destination for the budget that was bigger than the Marine Corps. And it has a name. The Defense Autonomous Warfare Group. DAWG, in Pentagon shorthand. The $54.6 billion office sitting inside Special Operations Command. The hidden destination for the budget that was bigger than the Marine Corps.
Reports tied to the FY2027 budget describe DAWG as the nucleus of a future autonomous warfare command structure. A peer to U.S. Cyber Command. Not officially a military branch. But increasingly functioning like one. Currently, DAWG operates inside SOCOM and draws on its special acquisition authorities. Its fast procurement pathways. Its political cover from the conventional services that would otherwise fight to grab the budget. A Southern Command version is already being built. The Southern Command Autonomous Warfare Command, or SAWC, was announced in early 2026. It will coordinate directly with DAWG on counter-narcotics drone operations in Latin America. Two regional autonomous warfare commands. Neither of them are legally a service. Both of them operate their own unmanned forces. The only thing missing is formal congressional recognition. And maybe that ambiguity is intentional. It is the seventh branch of the military in everything but name. A force without a fixed name is harder for Congress to scrutinize. Oversight becomes harder. It’s more difficult for foreign intelligence services to map. Russia and China both have spy services that work full-time mapping the American order of battle. They know how many F-35s the Air Force flies. They know how many submarines are tied up at Pearl Harbor on any given day. They have far less visibility into a force that has no building or service.
The fallout runs deeper than office paperwork. DAWG was assembled in months and funded by accounting categories most Americans have never heard of. It’s aimed at a war most Americans don't realize the Pentagon is actively preparing for. There are real consequences to building a force this way. The first is oversight. Congressional committees have established habits for scrutinizing the existing services. The House Armed Services Committee has subcommittees for tactical air, sea power, and strategic forces. There is no subcommittee for autonomous warfare. Senators Roger Wicker and Mike Rogers chair the House Armed Services Committee. Both have publicly cautioned against approving structural shifts of this size without clear oversight. None of them have stopped the budget request from advancing.
The second is execution. An office that obligated $225.9 million last year doesn’t have the contracting staff to spend $54.6 billion in 12 months. Replicator already missed its targets. Defense analysts predict DAWG will obligate less than 60% of its budget in the first year. The third is escalation. Carriers and bombers move slowly. They give diplomats time to negotiate. A swarm of pre-positioned autonomous systems is programmed to engage on its own once a threshold is crossed. Political leaders may find themselves in a war the machines started before diplomacy had a chance.
By the early 2030s, the Ford-class will still exist but its mission profile will look different. The carriers won't push into the Taiwan Strait. They will loiter outside the missile envelope, launching unmanned systems forward. They’ll serve as command nodes for swarms operating hundreds of miles away. The fighter pilots will still fly, but there will be fewer of them. Their job will be to manage formations of autonomous wingmen instead of flying individual jets into contested airspace. The Marines will still be active, but their landing forces will be smaller and more dispersed. They will be surrounded by autonomous reconnaissance and fires that sense and engage threats before any Marine sets foot on a beach.
But the center of gravity has shifted. The most expensive line in the Pentagon budget is no longer building a new fighter or a new carrier. It is building a swarm. The fight that everyone in Washington believes is coming is the fight over Taiwan. And it no longer believes the force structure built for the 20th century can win it. It will be won, or lost, by a force the Pentagon put together in plain sight. Behind a budget line nobody noticed. Inside a special operations command that never said out loud what it was being built to do. A military service in everything but name. The first shot of the next major-power war may be fired by a machine selecting targets faster than humans can intervene. Code will decide the outcome before humans even arrive. The seventh branch is already operational. It just hasn't been officially announced yet.
Thanks for joining me today. I hope you enjoyed the video, and if you want to see more videos like this, then please consider subscribing and turn on notifications. It really does make a difference. Thanks again, and I'll see you next time on Map Pack.