All podcasts / Invest Like the Best / Summary

Scott Nolan - SpaceX, Founders Fund, and Rebuilding American Uranium Enrichment

2026-04-14 - 78 min - source - Read full transcript

Patrick O'Shaughnessy (host)Scott Nolan

Key insights

Trends attract two layers of competition that compete away returns: company-level (new entrants chasing the same theme) and investor-level (capital bidding up price), so Founders Fund deliberately hunted for important, unworked problems instead.

Nolan frames this as the core reason Founders Fund avoided popular themes: even a good idea gets competed down to equilibrium returns once everyone piles in, on both the operating side and the pricing side.

contrarian-investing

Cheap valuations in venture are usually a red flag rather than a bargain, and investors compound the error by anchoring on the last round's price instead of the next one - so the steeper a company's up-round, the more undervalued it likely still is.

Nolan says a 'crazy value deal' in true venture investing empirically signals something wrong with a company's ability to raise capital at market price; he attributes the up-round heuristic to Peter Thiel, explaining that the only 'hard data' available is historical pricing, which biases people toward comparing against the past rather than forecasting the next markup.

contrarian-investing

An investor should not fall in love with an idea, but a founder has to be - the two roles require opposite emotional postures toward the same problem, and quality of life is structurally worse once you cross into operating.

Being in love with an idea as an investor causes compromises on team quality and throwing good money after bad; but starting a company is irrational enough that a founder needs deep personal conviction to sustain it, since operating problems are non-optional and immediate while an investor can always decline a meeting and only discover the cost of that pass years later.

founder-vs-investor-mindset

Founders Fund selects people who explicitly want to be investors, not frustrated operators, which is why the firm rarely spins out founders despite meeting thousands of companies.

Nolan describes his own jump to General Matter as a near-disservice-to-not-do situation: over a decade he met every advanced-reactor company, all of them cited the same missing input (fuel), and his SpaceX/Founders Fund background made him uniquely positioned to solve it - a case-by-case exception rather than a pattern.

founder-vs-investor-mindset

The U.S. has full domestic capability across the nuclear fuel supply chain except one step - enrichment - which is the actual bottleneck constraining all forms of nuclear energy, not reactor licensing or physics.

Of the five steps from mined ore to fuel pellet, only enrichment lacks commercial-scale U.S. capacity; every advanced-reactor company Nolan met cited the same missing input, which is what led him to conclude a new private company had to be built to solve it.

nuclear-fuel-bottleneck

Three sequential 'nuclear fuel cliffs' are forcing the issue: no reliable HALU supply for advanced reactors today, a full ban on Russian enriched uranium imports effective January 1, 2028 (removing roughly 25% of U.S. supply), and eventual depletion of the Navy's stockpile.

General Matter is targeting the HALU cliff first because it is the smallest, most urgent, least-contested market, then expanding into low-enriched uranium (LEU) for the existing 94-reactor grid, a roughly $2-2.5 billion U.S. market.

nuclear-fuel-bottleneck

Enrichment cost, not fuel availability, will ultimately determine whether advanced reactors succeed - fuel can be half of an advanced reactor's total operating cost, and enrichment is projected to be half of that fuel cost.

For gigawatt-scale light-water reactors, fuel cost is a small share of total cost (CAPEX dominates) so availability and supplier diversity matter more than price; for advanced reactors running on HALU, the enrichment cost is the dominant driver of their economics.

nuclear-fuel-bottleneck

General Matter's North Star metric, dollars per kilogram-SWU, is explicitly modeled on SpaceX's cost-per-kilogram-to-orbit, treating enrichment cost reduction as the single number that captures whether the business is achieving its mission.

Nolan draws a direct parallel between the two companies: SpaceX optimized cost per kilo to a given orbital velocity, General Matter optimizes cost per kilo of uranium to a given enrichment level (measured in separative work units).

nuclear-fuel-bottleneck

Vertically integrating engineering and manufacturing under one roof - rather than subcontracting through fixed interfaces - lets a hardware team trade off requirements in real time and avoids the calcified system architecture that results from deep subcontractor chains.

Nolan cites SpaceX-era aerospace projects with roughly 30 layers of subcontractors and a nuclear industry analyst call bragging about 900 subcontractors as examples of the pattern General Matter is deliberately avoiding by co-locating engineering with first-of-a-kind manufacturing, following the Tesla playbook of building an in-house EPC team rather than hiring a general contractor.

vertical-integration

Energy consumption per capita and GDP per capita are tightly correlated across essentially every country on Earth, yet U.S. grid capacity has been effectively flat since the 1990s while China has grown to roughly triple U.S. total energy production.

Nolan calls this the ultimate proxy for human prosperity and argues the U.S. cannot remain economically relevant without expanding energy production, especially as AI data-center demand threatens to consume the entire grid's growth by 2030 on current trajectories.

energy-and-economic-growth

Data centers can win faster permitting and community goodwill cheaply by 'bringing your own energy' (BYOE) - adding a small increment of net-positive power production alongside their own compute buildout rather than straining local grids.

Nolan estimates that a hyperscaler splitting capital 50/50 between compute and power could add roughly 100 extra megawatts to a community's grid for only a 5% project cost increase, which would meaningfully lower that community's utility rates and remove the 'will we be allowed to build here' risk that is now the biggest obstacle to speed-to-power.

energy-and-economic-growth

The winning argument for nuclear adoption is cost, not safety data - people don't find statistical safety arguments compelling, but a credible promise to cut utility bills in half changes minds.

Nolan argues nuclear already wins on being the safest and cleanest baseload source by the data, but that argument alone hasn't moved public opinion; the more persuasive pitch reframes nuclear around first-principles cost potential (a pellet of nuclear fuel is energy-equivalent to roughly a ton of coal) once enrichment and reactor costs are brought down.

energy-and-economic-growth

Companies

SpaceX - Nolan was employee number 35, working on Merlin engine systems and later the Dragon capsule under NASA's COTS program; Founders Fund's first hardware bet and the formative proof that unfashionable physical-world companies could win
Founders Fund - Nolan spent over a decade there as an investor, joining in 2011 under Peter Thiel's thesis of backing founder-controlled companies in underappreciated, non-digital ('atoms not bits') categories
General Matter - The company Nolan founded out of Founders Fund to rebuild domestic uranium enrichment capacity, starting with HALU for advanced reactors and later LEU for the existing grid
Boeing - Nolan interned there during college and cited it as an example of the stagnant, cost-plus incumbent aerospace industry that pushed him toward SpaceX
Square - Nolan nearly dropped out of Stanford business school in 2010 to join Square before being recruited by Peter Thiel into what became Founders Fund
Airbnb - Cited as a Founders Fund investment that succeeded outside the cost-plus-industry pattern; data on demographic shift and market-share gains convinced the firm the niche 'air mattress' business would go mainstream
Spotify - Cited as a Sean Parker-led investment shaped by his Napster-era understanding of music licensing and geography
Planet Labs - Nolan's first investment at Founders Fund, a satellite company
Crusoe Energy - A Founders Fund energy investment built around monetizing stranded power supply (e.g., flare gas)
Radiant - A microreactor company Founders Fund backed to serve stranded energy demand (e.g., remote communities, army bases); meeting reactor companies like this is what led Nolan to discover the enrichment bottleneck
Anduril - Cited alongside SpaceX and General Matter as an example of a company attacking a stagnant, cost-plus, oligopolistic incumbent industry (defense)
The Boring Company - Cited as another example of attacking a stagnant, cost-plus infrastructure industry
Tesla - Referenced as the playbook General Matter is following for in-house construction: build your own EPC (engineer-procure-construct) team rather than outsourcing to a general contractor
NASA - Nolan's SpaceX-era counterpart on the COTS program; described as full of true believers in space who collaborated openly with a credible private company

Techniques and frameworks

Avoid trends / avoid competition - Thiel's Founders Fund framework: a trend attracts both company-level competition (many entrants competing away returns) and investor-level competition (crowded capital bidding up price), so look for important problems nobody else is working on
Cost-plus stagnation pattern - Industries stuck in cost-plus contracting (aerospace, defense, infrastructure) lose all incentive to cut costs or innovate, calcifying into oligopoly - which is exactly where a disruptive new entrant can win big
Concentrate into winners, don't index - Founders Fund's approach to allocating capital: trust early intuition, avoid diluting returns by spreading bets, and be willing to double down on already-popular companies before they get even more popular
Steeper upround, greater undervaluation - Thiel's heuristic that investors anchor on the previous round's price rather than the next round's, so a company growing fast enough to post a large step-up is usually still undervalued relative to where it's going
Vertical integration / co-located engineering and manufacturing - Bringing subsystem design and manufacturing in-house (rather than layers of subcontractors with fixed interfaces) lets teams trade off requirements in real time and design for manufacturability, borrowed from SpaceX and applied at General Matter
BYOE (bring your own energy) - Nolan's concept for data centers: pair new compute buildout with a modest amount of extra power production that nets positive to the local grid, winning community goodwill and permitting speed cheaply
North Star metric: dollars per kilogram-SWU - General Matter's cost metric (dollars per kilogram of separative work unit), explicitly modeled on SpaceX's cost-per-kilogram-to-orbit as the single number that captures whether the business is winning
Elon algorithm (informal) - Nolan's description of SpaceX engineering culture: define the goal, avoid analysis paralysis, get to 90-95% and operational rather than 99% and stalled, then iterate

Summary

Scott Nolan's career traces a straight line from operator to investor and back to operator: employee number 35 at SpaceX working on engine and Dragon capsule systems, then more than a decade at Founders Fund investing alongside Peter Thiel, and now founder of General Matter, a company rebuilding domestic uranium enrichment. The throughline he offers for all three moves is the same question - what important problem is nobody else solving that I can actually contribute to - applied first as an engineer avoiding a stagnant incumbent aerospace industry, then as an investor hunting for underappreciated "atoms, not bits" categories, and finally as a founder confronting a bottleneck he could not find anyone else willing to fix.

The investing portion of the conversation is a tight articulation of the Founders Fund playbook: avoid trends because they invite both company-level competition (new entrants chasing the same theme) and investor-level competition (capital bidding up price), and instead look for stagnant, cost-plus, oligopolistic industries - aerospace, defense, infrastructure - where incumbents have no incentive to cut costs or innovate. Nolan pairs this with specific pricing heuristics: cheap valuations in true venture are usually a red flag rather than a bargain, and investors chronically anchor on the last round's price instead of the next one, so the steeper a company's up-round, the more undervalued it likely still is. He is explicit that an investor should never fall in love with an idea, even as he insists a founder has to be, because starting a company is irrational enough on pure economic terms that only deep personal conviction sustains it through a materially worse quality of life than investing offers.

The back half of the episode is a detailed walkthrough of the actual bottleneck General Matter exists to solve. The U.S. retains full domestic capability across the nuclear fuel supply chain except one step, enrichment, which every advanced-reactor company Nolan met during his Founders Fund years cited as their real constraint - not licensing, not physics. He lays out three sequential "nuclear fuel cliffs": no reliable supply of HALU (highly assayed low-enriched uranium) for advanced reactors today, a full ban on Russian enriched-uranium imports taking effect January 1, 2028 that removes roughly a quarter of current U.S. supply, and eventual depletion of the Navy's enriched-uranium stockpile. General Matter is targeting HALU first as the smallest, most urgent, least-contested market, then expanding into low-enriched uranium for the existing 94-reactor grid.

Operationally, Nolan describes building General Matter as an explicitly engineering-driven company modeled on SpaceX and Tesla: vertically integrate engineering and manufacturing under one roof rather than subcontracting through fixed interfaces (he cites aerospace projects with roughly 30 layers of subcontractors and a nuclear company bragging about 900 as the pattern to avoid), build an in-house construction team rather than hiring a general contractor, and track progress against a single North Star metric - dollars per kilogram-SWU - explicitly modeled on SpaceX's cost-per-kilogram-to-orbit discipline.

The conversation closes on a broader argument about energy and American competitiveness: energy consumption per capita correlates tightly with GDP per capita across nearly every country, yet U.S. grid capacity has been flat since the 1990s while China has grown to roughly triple U.S. total production. Nolan's proposed fix for AI-driven demand strain is "bring your own energy" - data centers pairing their compute buildout with a small increment of net-positive power production to win community goodwill and faster permitting - and he argues nuclear's path to real adoption runs through cost, not safety statistics: people are unmoved by risk data but respond to the concrete promise that nuclear could eventually cut their utility bill in half.

Notable Quotes

"It's what important problem is there that's not going to get solved otherwise that somehow I can contribute to." - Scott Nolan

"The steeper the upround, the greater the undervaluation." - Scott Nolan, attributing the line to Peter Thiel

"The team you build is the company you build." - Scott Nolan, citing Naval Ravikant, then adding: "the early team you build is the company you build"

"What if your utility bill got cut in half?" - Scott Nolan

"It's actually wrong to not go do that." - Scott Nolan, on why he left investing to found General Matter