“It turns out the primes are actually not helpful. Now, as a venture capitalist, I’m pattern matching for who is unwilling to partner with primes as a positive.”

Ouch.

Those words came from Trae Stephens, co-founder and chairman of Silicon Valley-based Anduril Industries and partner at Founders Fund, a venture capital firm investing in companies building “revolutionary technologies.” The comments came while speaking during a roundtable discussion Defense News hosted in California, when he pointed to two companies that have earned investors millions of dollars in returns off the Pentagon: SpaceX and Palantir.

Neither ever partnered with a prime, nor did they rely on small, one-off $250,000 contracts from some Defense Department pilot program that delivers no recurring revenue and therefore is (in the words of Stephens) meaningless.

Of course, both were founded by billionaires, so there’s that.

And we should probably acknowledge that the sentiment is not always much more cordial from the other side, traditional defense companies to Silicon Valley. In 2015, Northrop Grumman CEO Wes Bush offered a warning to the Pentagon about neglecting the defense industry as a key source of innovation for national security: “This is what we observe in the animal kingdom when resources become scarce. They begin feeding on each other.”

Again, ouch.

So what’s the problem? Are these two communities so drastically different that they can’t come together for the benefit of both sides?

Anyone could have seen potential for industry tensions. But when the Pentagon touted plans to cozy up more with Silicon Valley tech firms, the focus remained on challenges that permeated mostly from inside the building. Could the acquisition community adopt more commercial practices to speed up the process? Could watchdog agencies accept a degree of risk and even failure for the sake of innovation? Could program managers rationalize to Congress investment in new technologies, and perhaps let legacy systems officially die?

Those challenges still exist. But seemingly taken for granted was this notion that traditional defense companies and Silicon Valley-type tech firms could play nicely in the same sandbox. So now here we are.

This is not to say unions between Silicon Valley and the traditional defense companies don’t happen successfully.

Pivotal, also at the roundtable, teamed with Raytheon to help the Defense Department roll out the development of new platforms and transition legacy applications to cloud environments. That teaming remains intact after more than a year. And many smaller tech firms do rely on larger defense companies to take on the contracting morass so they don’t have to.

But to Stephens’ point, and as restated by not only Silicon Valley executives at the roundtable but also leadership from the Defense Innovation Unit and the Defense Innovation Board that also attended, technology is not the problem.

Likewise, to hand over products to ultimately become part of the Pentagon bureaucracy — without forcing defense agencies to take risks, to go out of their comfort zone, to change not just what they buy but how they buy and implement — defeats the purpose. In that case Wes Bush is exactly right — stick with the companies that have perfected the process and give them the credit that they deserve and the respect they’ve earned in knowing how to work within the machine.

Obviously, the challenges that face the Pentagon in successfully and consistently cooperating with Silicon Valley go well beyond the sometimes troubled relationship with traditional defense industry players. Cultural challenges abound.

But if these two communities are to successfully come together, industry players will need to learn to not just play nice, but play together.

In January, look for a special multimedia report that digs into the discussion from “Cultural Mesh, a DoD-Tech Dialogue.”