This interview features investor Scott Nolan, who realized the US has no domestic uranium enrichment capacity and relies on Russia for fuel for advanced nuclear reactors. He founded General Matter to rebuild it. He believes AI data centers will force tech companies to build their own nuclear plants, and fuel cost is key. Key holdings: SpaceX (he was employee #35, praises its vertical integration), Planet Labs (his first investment, a satellite company), and Radiant (a micro-reactor startup facing a fuel shortage).
After 12 years at Founders Fund, Scott Nolan founded General Matter with the goal of rebuilding U.S. uranium enrichment capabilities. The core thesis is that uranium enrichment is the sole bottleneck for America's nuclear future, and the window to address it is extremely narrow. The U.S. was once a
Scott Nolan, after 12 years as an investor at Founders Fund, founded General Matter with the mission to rebuild America’s uranium enrichment capabilities. The core thesis of this issue: Uranium enrichment is the sole bottleneck for America’s nuclear future, and the window to address it is extremely narrow. The United States was once the global leader in uranium enrichment, but completely halted such activities after the 1980s. Today, roughly a quarter of its enriched uranium is imported from Russia, and an import ban set to take full effect in 2028 leaves advanced reactors—touted as the next-generation power source for data centers—facing a crisis with no reliable domestic fuel supply. Nolan emphasizes that a passion for the problem itself is essential for founders, but for investors, falling in love with an idea is dangerous. Drawing on SpaceX’s experience with vertical integration, he argues that rebuilding this industrial capability, which has atrophied over 40 years, requires a fundamental shift in industrial strategy.
Scott Nolan believes that the core of both investing and entrepreneurship lies in identifying problems that are "important but unaddressed," and such problems are often hidden in industries that have stagnated.
During his 12 years at Founders Fund, Nolan's core strategy was to seek out areas that are "important but underinvested." He summarizes the typical characteristics of such problems: industries have fallen into a cost-plus oligopoly, lacking incentives for innovation. SpaceX (which he joined as employee No. 35) and many hardware companies backed by Founders Fund fit this pattern — before SpaceX, the space launch industry was dominated by a few giants surviving on government contracts, with no motivation to reduce costs.
He describes his method for identifying such opportunities: "When you meet a founder who talks about things no one else is really discussing, telling you why everyone who has thought about this problem either thinks it's impossible or is completely wrong — that conversation is usually not a surface-level pitch, but a deep dive down the rabbit hole." (Meaning: truly important opportunities often lie in the blind spots of mainstream consensus.)
Nolan began focusing on nuclear energy in 2010, investing in Planet Labs, Crusoe Energy, and microreactor company Radiant. Through Radiant, he discovered a common bottleneck across the entire advanced reactor industry: every reactor company told him that the biggest obstacle was not NRC licensing, but fuel — fuel came from Russia, and only Russia could produce it. He spent the entire 2023 studying the five stages of the nuclear fuel supply chain (mining → conversion → enrichment → reconversion → fuel rod fabrication), ultimately identifying enrichment as the sole bottleneck. Finding no company addressing this problem, he decided to found General Matter himself.
Nolan argues that the key to rebuilding America's industrial capacity is not simple "reshoring," but thorough vertical integration — placing engineering design and manufacturing under the same roof.
He uses SpaceX's experience as an example: during the Space Shuttle era, a system involved 30 layers of subcontractors, with each interface fixed and unoptimizable. SpaceX's approach, by contrast, is to "pull engineering together, vertically integrate, and take on more engineering work in-house rather than outsourcing it." This model allows the team to optimize across layers in each iteration — the mechanical team can tell the electrical team, "Your requirement is too tough for me; can you give me some margin? I'll make up for it elsewhere."
Nolan extends this philosophy to General Matter: "We even build our own factories — learned from Tesla, not hiring a general contractor to outsource everything, because then you lose control over the schedule." He emphasizes that for rebuilding an industrial capacity that has shrunk by 40%, the "learning curve" exists not only in manufacturing but also in the tight coupling between design and manufacturing. The U.S. outsourced a large amount of manufacturing overseas in the past, effectively outsourcing innovation as well — because iteration during the manufacturing process is the true source of innovation.
Nolan points out that the U.S. nuclear industry faces three "fuel cliffs," and General Matter is targeting the most urgent one first: HALEU (High-Assay Low-Enriched Uranium, approximately 20% enrichment) required for advanced reactors.
The nuclear fuel supply chain consists of five steps: mining → conversion to gas → enrichment → reconversion to solid → fabrication into fuel rods. The U.S. has capabilities in all steps except enrichment, where it is entirely absent. Nolan describes three critical junctures:
| Fuel Cliff | Type | Enrichment Level | Timeline | Impact |
|---|---|---|---|---|
| First Cliff | HALEU (Advanced Reactor Fuel) | ~20% | Next few years | All SMR companies lack a reliable fuel source, unable to scale |
| Second Cliff | LEU (Existing Reactor Fuel) | 3-5% | January 1, 2028 | Full ban on Russian imports takes effect, accounting for 20-25% of U.S. imports |
| Third Cliff | Naval Propulsion Enriched Uranium | High enrichment | Longer term | U.S. strategic reserves depleted |
General Matter's initial strategy is: serve the smallest market first — HALEU, because it is the most urgent niche with no competitors. This aligns with Founders Fund's classic strategy: "Start with a very small market, own it first, then grow into a larger one." In the second phase, the firm will enter the LEU market — the U.S. has 94 existing reactors, representing an annual market size of approximately $20-25 billion.
Nolan clarifies the relationship between enrichment and weapons-grade uranium: all enrichment is fundamentally the same, differing only in the degree of separation. Commercial-grade LEU is 3-5%, HALEU is 20% (the internationally recognized non-proliferation ceiling), and weapons-grade requires over 90%. He emphasizes that General Matter produces only commercial-grade products, and the international consensus is that all countries should remain below 20%.
Nolan argues that the explosive growth of AI data centers will force tech companies to "bring their own energy," and this could become a catalyst for a nuclear renaissance.
He highlights two key charts: one showing a strong correlation between GDP and per capita energy consumption (R²>0.8), and another indicating that the U.S. power grid has barely grown since the 1990s, while China's total energy output this year will be three times that of the U.S. He believes that if the U.S. does not proactively expand energy production, it risks a decline in economic competitiveness.
Nolan introduces the BYOE (Bring Your Own Energy) concept: Hyperscale data center operators, when building new data centers, will simultaneously invest in constructing their own power generation facilities (nuclear reactors or natural gas turbines) and reach agreements with local communities—not only avoiding drawing power from the grid but also supplying excess electricity back to the grid, thereby lowering electricity prices for local residents. He estimates that for a 1 GW data center, increasing generation capacity by 10% (adding only 5% to project costs) could provide an additional 100 MW of power to the community, "which would significantly reduce utility prices in that community."
He believes that for advanced reactors, fuel costs could account for more than half of their total power generation costs (in contrast, traditional large reactors have a very low fuel cost share, with the main cost being infrastructure). Therefore, General Matter's ability to reduce costs directly determines the economic viability of advanced reactors.
Nolan emphasizes that the early team determines a company's DNA, and General Matter is positioned as an "engineering-driven company," not a "science project."
He chose to locate the company in Southern California because nuclear engineers account for only a single-digit percentage of the team, while mechanical, electrical, software, and chemical engineers form the core — talent concentrated in Southern California's hardware/aerospace cluster. He personally conducts the final round of interviews, screening not only for skills but also for attitude: "They know there are many places to work, but this is the only private company working on this problem; they know there may be places with higher pay and easier work, but our mission requires delivery before the end of this decade — which means many late nights and weekends."
Nolan admits that an operator's quality of life is far inferior to an investor's: "Once you start a company, you put on the harness and have to deal with whatever comes your way. As an investor, you can choose not to meet with a certain project, but as an operator, when a problem arises, you have to solve it." However, he believes that when a problem is so important that "not doing it would be a dereliction of duty," starting a company is the only right choice.
| Position | Guest Stance | Key Data |
|---|---|---|
| SpaceX | Bullish (early employee, continuous investment) | Employee No. 35; Founders Fund first invested in 2008 |
| Planet Labs | Bullish (Nolan's first investment at Founders Fund) | Satellite company |
| Crusoe Energy | Bullish (Founders Fund investment) | Uses stranded natural gas to power data centers |
| Radiant | Bullish (Founders Fund investment) | Micro nuclear reactor company, targeting "isolated demand" for remote areas/military bases |
| Airbnb | Bullish (Founders Fund investment) | Early angel round + subsequent large rounds; data analysis revealed it was gaining share across markets |
| Spotify | Bullish (Founders Fund investment) | Led by Sean Parker, based on deep understanding of the music industry |
| Tesla | Not explicitly stated (used as methodological reference) | Vertical integration, in-house factory construction, etc., were referenced by General Matter |
1. "Avoiding trends" has two layers of competitive implications (Scott Nolan): At the company level—multiple companies compete within the same trend, driving profits toward the equilibrium of perfect competition; at the investor level—trends attract large amounts of capital, inflate valuations, and eliminate excess returns. The combination of these two layers makes trend investing a dangerous game.
2. "The last round's valuation is an anchoring bias; only the next round's valuation matters" (Scott Nolan, paraphrasing Peter Thiel): When a company experiences a significant valuation jump (e.g., 2x the previous round), it might actually deserve a 4x increase—people are merely anchored by historical prices. Investors should focus on "what catalysts can drive the next round higher" rather than "how much cheaper it is compared to the last round."
3. It is dangerous for investors to be obsessed with an idea; founders must be obsessed with it (Scott Nolan): Investor obsession with an idea can lead to compromises on the team and loss of objective judgment when making follow-on investments; whereas if a founder lacks sufficient passion, they cannot endure the irrational hardships of entrepreneurship. The two roles demand fundamentally different modes of emotional commitment.
4. The core advantage of vertical integration is cross-layer optimization (Scott Nolan): When engineering design and manufacturing are under the same roof, the mechanical team can tell the electrical team, "This requirement is too difficult; give me some margin, and I'll make up for it elsewhere"—such a conversation is nearly impossible between two independent companies. Fixed interfaces lead to rigid system architectures.
5. Nuclear renaissance faces three "fuel cliffs" (Scott Nolan): HALEU (for advanced reactors, in the coming years), LEU (for existing reactors, when the Russian ban takes effect in 2028), and naval propulsion (further out). General Matter enters from the most urgent and smallest HALEU market, then moves into the LEU market.
6. BYOE (Bring Your Own Energy) will become the standard for data center construction (Scott Nolan): Hyperscale data center operators will simultaneously invest in power generation facilities, not only avoiding strain on the grid but also supplying surplus electricity to the community and lowering residential electricity rates. For operators, the speed advantage of "being accepted by the community" far outweighs the additional cost of power generation.
7. Fuel costs for advanced reactors may account for more than half of their total power generation costs (Scott Nolan): In contrast, fuel costs for traditional large reactors are very low, with the main cost being infrastructure. This means General Matter's cost-reduction capability directly determines the economic viability of advanced reactors—"If nuclear power is cheaper than natural gas, people will suddenly find it very attractive."
8. Operator and investor are two fundamentally different lives (Scott Nolan): An investor can "choose not to meet with that project" to protect quality of life, but an operator "must deal with problems when they arise." However, when a problem is so important that "not doing it would be a dereliction of duty," starting a company is the only correct choice—even if it means a decline in quality of life.