The future of war has arrived in Ukraine. That country’s defenders are able to hold back a Russian advance, even though the Russians have a manpower advantage of as much as 5-to-1 along some parts of the front line, largely by using drones. By some estimates, unmanned aerial systems are now inflicting 70 percent of all casualties on both sides, reducing traditional weapons such as tanks and artillery almost to irrelevance. The war has also ushered in the use of ground-based and sea-based drones — indeed, using the latter, Ukraine managed to defeat Russia’s Black Sea Fleet.

The drone revolution necessitates an urgent effort by the U.S. military to catch up, but instead of looking to the 2020s and beyond, the commander in chief has his eyes firmly fixed on the 1980s.

President Donald Trump came to office with dreams of reviving President Ronald Reagan’s plans for using space-based interceptors to protect the United States from nuclear missile attacks. He initially called his blueprint Iron Dome for America, after one of the systems that help protect Israel (a country the size of New Jersey) from missile attack. Now, in keeping with Trump’s fixation with gold (the dominant motif in the redecorated Oval Office), it has been renamed Golden Dome.

Last week, Trump unveiled more details about Golden Dome during an Oval Office event, and named a manager (U.S. Space Force Gen. Michael Guetlein) for the project. Trump claims Golden Dome will cost $175 billion to build; protect America not only against intercontinental ballistic missiles but also hypersonic missiles, cruise missiles and drones; and be operational by the time he leaves office in 2029.

And, if you believe that, the president has some memecoins he would be happy to sell you. In reality, the cost of developing and operating space-based interceptors is estimated by the Congressional Budget Office to be as high as $542 billion over the next 20 years, and missile-defense experts say it will take at least 10 years for such a system to be operational — if it’s possible at all. And, even if such a system is deployed, it is very unlikely to provide an effective defense for the entire country against nuclear attack.

As Michael O’Hanlon of the Brookings Institution argues in the Wall Street Journal, there is a case for a less ambitious missile-defense system augmenting and linking together existing capabilities such as the Terminal High Altitude Area Defense (THAAD) and the Navy’s Aegis Combat System. But that would still be mainly a defense against a small-scale attack from a rogue state such as North Korea or Iran. A space-based interceptor system to protect the United States against a massive nuclear attack from Russia or China remains as much of a fantasy today as when Reagan first proposed it in 1983.

Instead of devoting defense dollars to Golden Dome, the Trump administration would be better advised to focus on building lots and lots of cheap drones. The Defense Department estimates that the United States has the capacity to manufacture 100,000 drones a year.

That sounds like a lot, but actually it’s a pittance. Last year Ukraine produced 2.2 million drones, and this year it’s aiming to build 4.5 million.

Admittedly, most of the Ukrainian drones are cheap, first-person-view models that are far less sophisticated than high-end drones like the $28 million MQ-9 Reaper or the $140 million RQ-4 Global Hawk that defense contractors sell to the Pentagon. But the Russia-Ukraine conflict has shown the value of cheap, disposable drones. They now dominate the battlefield. By contrast, bigger, more sophisticated drones may be too vulnerable to modern air defenses to be cost-effective: The Houthis shot down seven Reapers in less than six weeks over Yemen, costing the United States $200 million worth of aircraft.

Unfortunately, the United States lacks the capacity to build small drones in large numbers, even though U.S. scientists were pioneers in developing unmanned systems in the first place. The Chinese government used state subsidies to help build up a massive drone-manufacturing industry as part of its Made in China 2025 initiative. China was able to produce high-quality drones so cheaply that it made it impossible for manufacturers in the United States and other countries to compete.

Now just one Chinese company, DJI, is estimated to have 90 percent of the global drone market. Most of the drones that Ukraine makes employ Chinese-made parts, but obviously the Defense Department cannot be reliant for drone parts from a country that might become a wartime adversary. (Indeed, the Defense Department is forbidden under U.S. law from buying drones that have any electronic components made in China.) Last year, the House passed legislation, known as the Countering CCP Drones Act, to forbid DJI drone sales in the United States. But the bill did not advance in the Senate, in part because there is no viable, cost-effective, domestically produced alternative to DJI drones on the commercial market.

That needs to change, and fast. Congress and the administration need to recognize the strategic necessity of America developing a large drone industry of its own — one that cannot only supply U.S. consumers and companies but also the U.S. armed forces. The good news is that making drones isn’t that expensive, and it can utilize America’s high-tech know-how.

The Trump administration defense budget calls for spending $25 billion on Golden Dome this year. Imagine how far that money could go if the United States were to invest it in drone production. Kyiv is allocating $2.6 billion this year to build 4.5 million first-person-view drones at an average cost of just $580 apiece. If the United States were able to build drones as quickly and cheaply as Ukraine, it should be possible to build 43 million drones for $25 billion.

If the United States could produce 43 million drones a year, that would create a far more effective deterrent to Chinese aggression than any amount of investment in the unproven technologies of Golden Dome. The Chinese armed forces, which are buying large numbers of their own drones (actual numbers are secret), would know that they could never invade and occupy Taiwan in the face of such massive drone swarms.

Of course, it will never be as cheap to build drones in the United States as it is in Ukraine, because of the higher wages of U.S. workers — and the stultifying effect of the Defense Department’s procurement bureaucracy. But Ukraine is discussing the possibility of exporting drones to raise money for its war effort. The Trump administration could defend both the United States and Ukraine by investing in Ukraine’s drone industry and leveraging Ukrainian expertise to help build drones in America. As Post columnist Rahm Emanuel has argued, America needs drone technology from Ukraine far more than its mineral wealth.

A possible model is the 2022 Chips and Science Act passed during the Biden administration to revitalize America’s semiconductor industry.

Drones, like microchips, are a strategic technology that deserves federal subsidies — and substantial procurement from the Pentagon. That would be a far more effective investment than the Golden Dome boondoggle.

Max Boot is a Washington Post columnist and a senior fellow at the Council on Foreign Relations.