Trae Stephens - Find Good Quests - [Invest Like the Best, EP.319]
What this covers
Patrick talks with Trae Stephens, a Founders Fund partner and the co-founder and executive chairman of Anduril. The conversation moves between investing philosophy and the mechanics of selling to the Pentagon. Stephens begins with the claim that many high-margin software businesses (advertising, attention products, enterprise dev tools) pull engineers away from harder problems. He carries that into energy, arguing the US lost Eisenhower's cheap-energy vision after Three Mile Island and Chernobyl and leaned back into carbon.
The defense material is the longest stretch. Stephens traces the primes' stagnation to the 1994 "Last Supper" consolidation and to programs like the F-35, spread across roughly 400 congressional districts so they are nearly impossible to kill. He says the primes respond to their incentives rather than act with malintent, and he contrasts today's incentives with the speed-oriented Skunk Works era. He defends lobbyists as a necessary information layer for generalist legislators. On selling, he says pilots are easy to win but scaling them is hard, so companies should aim at core multibillion-dollar programs. He also makes the contested argument that nuclear deterrence has not stopped lower-level conflict, and he describes cost deterrence using cheap drones and expensive interceptors.
The investing thread is more personal. Stephens draws on Girard's mimetic theory to argue that copied ideas turn competition into a contest over capital. He opens pitches with the Origin Story Test, warns about mortgaging a company on 2021-era valuations, and values a fund's operational manager as highly as its partners. Smaller turns include Palmer Luckey on the physics limits of VR and a plea to cultivate "spiky" people over well-rounded ones.
Stephens argues that the most consequential companies pursue mission-driven 'good quests' in hard, strategically important sectors such as defense, because high-margin commercial software distracts talent and the defense procurement system has become too incumbent-protective to produce innovation without new entrants and aggressive top-down and bottom-up selling.
- High-margin software pulls engineers away from energy, hardware and defense problems that matter more, even though those businesses are lower margin.
- Defense primes ossified because consolidation and distributed contract incentives rewarded stability, so new entrants must break in through top-down and bottom-up selling.
- Founders need paranoia, fundraising skill and mission orientation, and investors should back them even at uncomfortable prices while managing capital strategy.
This asset isn't compiled yet
You're seeing its claims, ranked. Compile it to build the argument threads, weight them, and check each claim against your library — the full view.
René Girard's mimetic theory holds that most conflict arises from similarity and competition over scarce resources, because people want what others want, and the internet amplifies this by exposing people to what others are doing.
“mimetic theory, which is the idea that most conflict doesn't arrive from differences, it arises out of similarities, competition over scarce resources.”
Following René Girard's mimetic theory, people want what others want, which makes competition over copied ideas such as ridesharing become a competition over capital rather than technology, and in perfectly competitive markets capital is competed away so no one can become a large company.
“It becomes more of a competition over capital than a competition over technology.”
In the late 2020 and 2021 boom, high valuations, including 1,000x revenue multiples, led to down-round reissuance of shares when the market turned, which sapped morale, so founders should evaluate whether they are mortgaging the business on a high price.
“they had to reissue shares at a lower price. I mean this whole drama that really sapped the morale from the company.”
Palmer Luckey told Stephens that many VR aspirations, including optics and omnidirectional treadmills, are limited by basic physics that requires more science research before general-purpose VR like Ready Player One is possible.
“things like optics and omnidirectional treadmills and full immersion, all of this stuff is actually physics limited.”
Most technology that matters most to the Department of Defense over the next 20 years will be commercial, built by the private sector, so the DoD needs new entrants with access to talent and new technology, yet the procurement system is still built to exclude them.
“Most of the technology that matters the most to the Department of Defense over the next 20 years are commercial technologies”
Nuclear deterrence no longer works as it did in the Cold War because it requires a credible threat that someone will actually act, and tolerance for conflict such as in Ukraine, with over 100,000 deaths but no talk of strategic nuclear use, shows that US nuclear deterrence has not stopped lower-level conflict.
“over 100,000 people have died in Ukraine and nobody is talking about launching a strategic nuclear weapon.”
The 1994 Last Supper meeting, in which the Secretary of Defense told large contractors to consolidate or die, produced a few competitors per platform, but the approach distributed work across congressional districts and came at the expense of dynamism and new entrants.
“But it did it at the expense of dynamism. t did it at the expense of allowing for new entrants to come out of nowhere.”
Platforms such as the F-35 are built across roughly 400 congressional districts so that almost every member of Congress has jobs dependent on the program, making it nearly impossible to kill.
“platforms like the F-35 are currently made in 400 congressional districts to make sure that it's basically impossible to kill”
The large defense primes lost their innovative capacity not because of malintent but because they behave like government agencies responding to the incentive structures they are given, and the Cold War incentive structure was very different from today's.
“you have to think of the primes as basically, they're like government agencies, and they just respond to the incentives”
During the Cold War the defense primes operated under a different incentive structure that prioritized speed, so Skunk Works engineers iterated constantly on aircraft, and the primes respond to the incentives of their structure rather than harboring malintent.
“they're like government agencies, and they just respond to the incentives that they're given in the structure”
Lobbyists are a necessary information layer because modern members of Congress, who are mostly career politicians without prior expertise, cannot master the many topics they vote on, so lobbyists consume and communicate information to committee offices, though bad actors exist.
“this is effectively what lobbyists do in our system. They consume information, they communicate that in their network”
The US lost track of Eisenhower's Atoms for Peace vision of very cheap energy after nuclear accidents such as Three Mile Island and Chernobyl, and the net result was leaning back into the carbon economy.
“And then we just decided that it wasn't worth it anymore. There were, I guess, reasons related to nuclear accidents”
Many promised virtual reality features such as optics and omnidirectional treadmills are physics-limited and need basic research, but single-purpose devices like a reading headset can be built well today, in the same way that The Jetsons imagined specialized gadgets rather than one general device.
“most of the stuff that we talk about as being the future of virtual reality, things like optics and omnidirectional treadmills and full immersion, all of this stuff is actually physics limited.”
As a society we need to better cultivate and encourage people to be spiky, with deep specialized passions, rather than being well-rounded generalists.
“we need to do a better job of cultivating and encouraging people to be really spiky rather than being really well rounded.”
Modern warfare features low single-digit million dollar attributable drones that must be shot down with interceptor missiles costing over two million dollars each, making the trade economically irrational for the defender.
“drones that are in the low single-digit million dollar range. And in order to take them down, we have to shoot a two-plus million missile.”
Companies become monopolies because someone had an idea that no one else had, so investors should look for natural monopolies and contrarian ideas rather than people moving with the crowd.
“The reason that these things become monopolies is that someone had an idea that someone else didn't have.”
It has never been easier to win a government pilot, with programs such as SBIR, DIU, In-Q-Tel and over 100 DoD innovation organizations, but transitioning pilots into scaled programs is still very difficult, so companies should target core multibillion-dollar programs rather than small non-dilutive grants.
“It's never been easier than at any point in American history to win a pilot with the government.”
Many high-margin software businesses are advertising, attention-pulling or enterprise dev-tool companies, and they are bad for society because they draw engineers away from more important work, even though the software itself is not necessarily bad.
“they draw engineers in the tech community away from things that have more importance”
When pressed, most technology founders would admit their real mission is to make a lot of money, whereas founders in strategically important sectors are motivated by something bigger than themselves, which gives them a power the tech industry is lacking.
“most of them at the end of the day, it would just be making money.”
In 2017 Stephens argued that the Pentagon's core problems required a 21st century reboot of Lockheed Martin, a software-focused prime contractor with ambitious low-cost systems that could lower defense spending while increasing capability, which became the basis for Anduril.
“what we really need is a 21st century reboot of Lockheed Martin.”
Deterrence outside the nuclear regime depends on cost: if an adversary can press conflict cheaply while defenders must counter at much higher cost, the defender's deterrent edge breaks down, so Anduril aims to make aggression unthinkable by lowering the cost of countering it.
“We have a cost deterrence, can we counter someone's offensive in a way that's significantly more expensive for them than it is for us.”
Winning government contracts requires both user-level belief that the product works and top-down alignment with budget appropriators and decision makers who need to derisk the decision, then breaking through the risk-averse frozen middle who sign the paperwork.
“As you hinted, our strategy is top-down and bottoms up.”
The first question Stephens asks in pitch meetings is about the origin story, because founders who started from a whiteboard brainstorm for a business they can make money from are not disagreeable, whereas disagreeable founders believe something most peers would dispute and risk their reputation to prove it.
“The first question I generally ask in every pitch meeting is what is the origin story of this business.”
The quality of the fund's operational manager, such as Lauren Gross at Founders Fund, matters as much as the quality of the general partners, because she adds stability and relationship management that even a strong partner like Peter Thiel might not provide.
“I think the quality of your fund manager is also really, really important.”
Palantir's early strategy of partnering with consulting firms, high-channel resellers, and SAP to reach government customers did not work, because those partners had already sold to government and did not deliver the needed product.
“As it turns out, like, none of these things worked.”
Stephens argues that relativism, which treats each person's truth as just perspective, leaves people lonely and empty and lacks the foundation needed for community and belonging, while belief in objective truth, right and wrong creates meaning and optimism.
“my pushback to this is always the emptiness that you get out of relativism it drains your soul of all of your energy.”
Palantir's early channel partner strategy failed, and the company survived because it was deeply paranoid about deals being scooped by primes or bespoke solutions, and Shyam Sankar institutionalized that paranoia into organizational antibodies that pushed the team into hard customer engagements.
“And the reason that Palantir survived was really paranoia.”
Instead of general-purpose VR that is physics-limited by optics and treadmills, the near-term opportunity is single-purpose devices; Stephens co-founded Sol to build a lightweight, E-ink head-mounted display for reading books hands-free without notifications.
“This is a single-purpose device for people who want to read, people that don't want to be distracted with notifications”
A liberal arts mindset that tells people not to over-specialize leaves 18-year-olds facing an identity crisis when they must choose a career path in college, which leads to tracked, default decisions and regret by their mid-30s.
“they look back and they're like, I don't even know how I ended up here.”
Thiel observed that the founders of the most impactful technology companies are slightly socially off and not high in EQ, likely because they lacked the mimetic contagion that drives others to chase social validation, so they pursued unpopular things.
“a lot of the founders of the most impactful companies ... are slightly off socially”
Stephens wants to see founders who are paranoid and recognize that their problems are crazy hard with powerful incumbents, rather than overconfident founders who think success will be easy.
“I want to see them paranoid. I don't want to see them overly confident”
Building a product or investment strategy around the seven deadly sins is good for business but feeds humanity's lowest instincts, which pulls people away from their highest calling, so Stephens rejects it, arguing that drug tolerance in San Francisco is not love.
“that's feeding the lowest base part of humanity. And I don't think that's good for any of us.”
Six years at Palantir in a supporting role provided training in tech, software companies and entrepreneurship that Stephens says is roughly 100 times better than graduate school, and it also built relationships that feed his current investing and co-founding work.
“It's like 100x better than graduate school, living and breathing that supporting role.”
The most successful Founders Fund portfolio founders had low ego about the tasks they had to do and were willing to take on any role needed, which Stephens treats as a key indicator of success more than frenetic hours.
“having a willingness to actually just do whatever role needs to be done at the company, no matter how high status or low status it is”
Stephens believes each person has a sacred vocation, in which their giftedness is connected to their calling, so that profession, personal life and religious life are unified, and the pursuit of a calling makes life a religious experience that pushes the future toward a techno-optimistic outcome.
“we each have what I would call like a sacred vocation.”
Silicon Valley's meditation retreats and ayahuasca offer a momentary sense of spiritual connection, but without a foundation of objective truth these experiences leave people lonely, frustrated and empty, since personal truth collapses under conflict.
“your truth breaks down when your experience has changed and when it's in conflict with everyone else's truth”
The good quests idea, developed with Markie Wagner in a Pirate Wires essay, argues that successful people in a second act should focus on meaningful missions such as semiconductors, energy or defense tech, rather than commodity pursuits like lifestyle brands, celebrity businesses or generic VC and SaaS.
“how do you get the right people to focus on the right things in their second act?”
The most important work, involving critical natural resources and hardware, is much lower margin, and high-margin software changes the economic curve so these lower-margin but more important businesses are harder to start and harder to build network effects in, though they are no less important.
“they involve critical natural resources. They involve different forms of hardware. And so I think it changes the curve.”
Gov-tech entrepreneurs wrongly believe that if they build a great product the customer will come; in government the customer will not come unless the startup creates a strategy to ensure the product is actually used as intended.
“It is not the field of dreams. It's like the anti-field of dreams.”
At least one founding team member must be very good at fundraising, because founders who blame unenlightened VCs for failing to fund them absolve themselves of responsibility, and great storytellers like Elizabeth Holmes or Adam Neumann raised large sums regardless of business quality.
“fundraising is a skill and entrepreneurs need to be good at this skill.”
Ben Rich's Skunk Works experience involved constant iteration on many aircraft, but an engineer who joined Lockheed Martin in the early 1990s may have worked on only one or few, showing that development pacing has significantly slowed.
“The pacing has just gone significantly down.”
After joining Founders Fund in 2014 and meeting hundreds of government-focused founders, the firm made only one defense-related investment, Xpanse, which was later acquired by Palo Alto Networks, because the niche point solutions lacked scale.
“We ended up making one investment in the category, this company called that later was renamed Xpanse”
Companies selling into government should hire a lobbyist early, since it costs $100,000 or more per year, and Anduril began lobbying in its first week, briefing Capitol Hill offices within two days of founding.
“I think earlier is better. It's not cheap. I mean, you're going to spend $100,000 or more a year.”
The rule for investors remains that they should invest in rare founders in important sectors even at an uncomfortable price, while ensuring those founders manage capital strategy so that the company is not harmed by financing choices rather than operational failure.
“in almost every case, especially early, you should be deploying in even if the price is a little uncomfortable.”
If humanity had unlimited zero-emission energy, possibly through nuclear advances or a Dyson sphere, it would enable every other industry in ways that are hard to imagine, which is why he would snap energy into existence to most advance human progress.
“I think it has to be energy. What we could do as a planet if we had unlimited zero emissions energy”
Katherine Boyle's term American Dynamism captures investing in strategically important sectors, but the American or nationalist part matters less than the dynamism, meaning the pursuit of a techno-optimistic future through mission-driven founders.
“The American part of it is probably less interesting than the Dynamism part of it.”
Former Lockheed Martin CEO Norm Augustine joked that by 2050 the entire DoD budget would be spent on buying one airplane, a joke that Stephens says is not far from the trajectory of rising costs and falling quantities.
“in the year 2050, the entire DoD budget will be used to buy one airplane.”
Shyam Sankar's X-Men analogy holds that a company works when each member leverages their own superpower toward an audacious goal, and Stephens applies this to matching unique talent with lanes in hard technical fields he does not personally understand.
“Everybody has their own lane, and they're just doing the thing that is in their superpower lane.”
Tolerating open drug use on San Francisco streets, under the view of not interfering with people's lives, is not love; Stephens calls it a moral shortcoming of relativistic society.
“I don't think that's love at all.”
People make tracked decisions by doing the easiest or most natural next step, pushing choices further down a path, so by their mid-30s they cannot explain how they arrived where they are.
“People just do the thing that feels the easiest or the natural next step.”
After being rejected by almost every college he applied to, Stephens's mother told him to fly to Georgetown and tell the dean he would attend, an action that changed the trajectory of his life and taught him that the story is not written yet.
“the story is not written yet. You have the ability to impact the final story.”
Virtue is dismissed as an old-fashioned relic even though it appears in every culture and is critical to happiness, and humanity loses much of its ethical grounding when people treat themselves as the sole source of truth.
“Virtue is critical to happiness, to people's experience of the world.”
Some startups are founded in a whiteboard way, where people brainstorm and pick the least bad idea because their real goal is to be entrepreneurs and make money, and the idea is secondary.
“A couple of people standing in front of a whiteboard. They write every idea they have on the board, and then they narrow it down to the least bad idea.”
To help children develop passions, parents should find what excites and energizes each child, then give them opportunities to explore it and overinvest in chasing the passion to its conclusion while watching for burnout.
“I just want to like overinvest, making sure that they chase that passion, all the way to its conclusion.”
Stephens joined Palantir after seeing a demo, was recruited early, found he was not as technical as the government believed, and instead did sales, where he got to know Peter Thiel before Thiel recruited him to Founders Fund in 2013.
“I did sales at Palantir instead.”
Ryan Peterson of Flexport was obviously the right person to build the company from the first ten minutes of conversation, because he had exported ATVs and bath fittings from China, learned Mandarin, and built ImportGenius, showing an intense passion for freight logistics.
“There is not a better person in the world to build Flexport than Ryan Peterson”
As a parent, Stephens lets his children's demonstrated interests guide where he invests effort, overinvesting in whichever passion holds attention, which for his nine-year-old is creative pursuits and for his other child is physical activity.
“we try to direct him in the channels where his attention is held and where he is excited.”
Successful people who start lifestyle brands or Silicon Valley VC and SaaS businesses on the back of their brand are building commodities that do not advance humanity, which Stephens calls pathetic.
“It's a commodity. They're just building the same thing that everyone else has built”
Patrick observes that objectively successful people frequently report regret about a lack of virtue in their lives, and he suggests that slipstreaming virtues into people's thinking would be productive.
“effectively cite a lack of virtue in different clothing in different ways in their life and wishing that they could go back”
Since the 1990s Northrop Grumman has resembled a serial M&A and share-buyback story, showing strong capital returns in quantitative screens while the innovation story of the golden era of defense primes has faded.
“And then you look at Grumman since the '90s, and it looks like a Warren Buffett story or something. It was like a serial M&A story”
A Dyson sphere is not close to being built and humanity cannot yet imagine how to get there, but a Manhattan Project-level effort to reinvest in transformative energy research would be valuable.
“Are we close to actually building a Dyson sphere? No, of course, not.”
Stephens was aiming to be a foreign correspondent, but after 9/11 his motivation shifted from journalism to civil service, leading him to foreign policy and security studies at Georgetown and then government work.
“It was no longer about journalism it was about civil service.”
Breaking the loop of incumbency in defense procurement will require a very charismatic, heroic champion to reset the culture so the procurement system can work the way it needs to for future conflict.
“it will require a very charismatic heroic champion to come in and reset that culture”
Anduril should not build products in defense stack areas where many venture-backed companies already operate, because competing head-to-head there wastes resources.
“I don't think we should be building products that have a bunch of venture-backed companies already working in those spaces.”
The View reading device is a lightweight, Ray-Ban-sized headset with low-power e-ink that lets users read books hands-free without notifications, designed for people who want to read without distraction.
“This is a single-purpose device for people who want to read, people that don't want to be distracted with notifications”
Palmer Luckey, who was going to be fired from Oculus, agreed to co-found Anduril, and the company was launched in earnest in early 2017 after he told Stephens it was on.
“he said, I'm probably going to get fired soon. So as soon as I get fired, let's meet up”