Silicon Valley has been trying to shake up defense contracting for years. With Trump, they have a willing audience

• While Silicon Valley and Washington build a closer touchTech leaders offer advice on how the government can change better and faster. Founders and investors of tech tech startups said the Pentagon should reduce to lead hours and raise risk allowing levels to produce new weapons.

After years of trying to make incoming into the well -known Byzantine defense sector of the US government, the Silicon Valley is finally getting the opportunity.

A harvest of new defense startups from the valley goes to Washington at a time when the Pentagon is eager for the new tech. There are many leading numbers from tech reelection supported by President Donald Trump, who cement a new bond between an industry previously known for supporting democrats.

A recent conference on the country's capital has highlights a new close relationship between tech and government. The Hill and Valley Forum on Wednesday featured top defense tech firms such as Palantir's Alex Karp and Anduril's Brian Schimpf, who was washing shoulders with government officials such as then-national security advisor Mike Waltz as well as members of the Senate Armed Services Committee such as Joni Ernst (R-Iowa) (Rs.D.), and Jack Reed (Dr.I.).

Against the rear of deepening US geopolitical rivals in China, Tech leaders are asking for the government to take A page from its playbook found a happy audience.

The White House is “fully dedicated to reforming the way technology is taken” to modernize the US military, Waltz said the day before he left his role as a national security advisor.

Trump signed a lot Executive orders That will be a stronger of how the defense department gets new defense systems. Defense Tech Startups have long maintained that current methods have left them without competing with existing military contractors that they have viewed as having lower products but deeper relationships with Pentagon.

Executive orders are “going after things that always seem to be excessive, deliver too little and too long,” Waltz told the audience during a panel titled title Arsenal Reimagined: DOD designing for the 21st century of Battlefield. “We can fill this auditorium with defense and obtaining reform thinking of tank pieces, but you have a president and you have a leadership group that all gas, no brakes, and sometimes we can help them guide.”

In the midst of the conversations are the tendency of the Pentagon for long, expanded bidding processes and research projects, and a culture at risk that is more difficult for the DOD to have a tech experimental opportunity.

“There is a basic fact that the change is messy and chaotic,” said Palantir technology official Shyam Sankar.

On Friday, the White House submitted a 2026 federal budget included $ 1.01 trillion In funding for DOD. Defense Tech startups find themselves in a strange position of both failed in DOD operations, which they view as stodgy and anti-Meritocratic, and, at the same time focusing its business. Now, because of Silicon Valley's close relationship with the Trump administration, political allies appear to be found for the reforms he is looking for.

'You still shy'

But even if the DOD opens the process of getting it to tech companies and startups, they will still face a difficult market, according to Palantir's Karp.

“You still have to go upwards, but shooting up and shooting that Everest wants to mount Everest as they drop grenades with you is a different story,” said Karp, whose company has successfully charged the US army in 2016 for its blocking from bidding for a government contract. That move was widely considered to open Pentagon doors to the Silicon Valley.

Anduril's Schimpf suggested that the Pentagon should place large orders with defense startups. “If you buy things, the capital flows into defense,” he said. “Buy things on the size object, to move the needle and create opportunities to actually ride.”

Without the guarantees of large contracts, Anduril is “written only” that generates new versions of products such as air-to-air missiles that do not believe that a consumer can find, Schimpf added. “I don't think for 20 years anyone will buy any of the air-to-air missile we have made, because they are already focusing” on buying from other people, he said.

Emil Michael, Trump's nominee for Undersecretary of Defense for Research and Engineering, believes the Pentagon may be less dependent on specialized defense systems and is more open to existing commercial products when looking for a new tech to buy. “We don't need things that are always bespoke,” he said.

Michael, who has not been confirmed for his role in Pentagon, said the DOD said it could also benefit from looking at opportunities to save time, not just money. “Saving time is not something inherent in the DOD business model, [which is] About reducing the risk to the smallest possible substance at the cost of moving as quickly as possible. “

Quickly fail, fail often

In discussions about the development of new technologies, the conversation often turns to one of the Silicon Valley mantras: Fail quickly, not often fail. The idea, which is a tech culture element, is that the many failed iterations of a product are not important as long as the final version works.

“Failure doesn't matter. It's the size of success that matters,” venture capitalist Vinod Khosla said when asked about how to make the government comfortable getting the risk.

Palantir's Sankar suggested increasing the competition between the employees of the Defense Department to create, so they will have a “incentive to defeat the bureaucrat two doors to the corridor.” He considers the DOD to become a monopoly that has settled a change by being a single buyer of market defense systems.

Instead, Sankar suggested that many program managers have assigned the administration of the same project, with the contract eventually going to one that delivered a better outcome. “They will wake up every day like hyper-competitive Americans who try to kill each other,” he said. “There will be an incentive like 'yeah let faster faster faster. Let's do it better.'”

Speakers at the conference said that the continued geopolitical and Ai Arm breeding and China breed only added a more rush to the issue.

“And when you are in an AI breed when every change can lead to ten -ten -billion -billions, if not a way -a billion -billions, worth creating value -and you think of creating value as a better defense, shield, more restraint -every minute you lose is expensive,” Michael said.

This story was originally featured on Fortune.com