Simon Gros

---

At Harvard’s John F. Kennedy Jr. Forum, the Institute of Politics and Harvard’s Program on Constitutional Government convened a public conversation in which historian Niall Ferguson moderated a wide-ranging discussion with technology entrepreneur and investor Peter Thiel. The exchange unfolded as a structured interview followed by audience questions, moving across the near-term condition of Silicon Valley, the longer-run tempo of technological change, and the political economy of innovation in advanced societies.

Ferguson opened by pressing on a then-salient concern: whether the contemporary technology sector resembled a second dot-com bubble. Thiel’s answer was guarded but comparatively stabilizing. He argued that the late-1990s episode had been characterized by extreme valuations and a generalized exuberance that treated speed and scale as substitutes for durable business formation. By contrast, he described the current moment as one in which companies were being built more incrementally, with valuations less detached from underlying revenues and capabilities. Yet he paired this relative reassurance with a broader qualification: even if specific pockets of computing and internet activity were thriving, their aggregate impact remained insufficient to transform the overall economic trajectory, including in California itself. In this framing, localized technological dynamism coexisted with macroeconomic underperformance, and the coexistence itself became an empirical clue that the digital sector, for all its visible success, was not functioning as a comprehensive growth engine for society at large.

From there, Thiel addressed a recurring question directed at prominent investors: how to identify “the next” dominant company. He resisted promotional talk and instead used the prompt to distinguish two kinds of progress. He contrasted globalization—described as “horizontal” growth in which developing economies replicate existing practices—with technological innovation conceived as “vertical” progress in which new capabilities are created that did not previously exist. The former, he argued, can be approached mechanistically: one can observe what works elsewhere and copy it. The latter, by definition, is non-repeatable in its decisive features: major innovations occur in particular contexts and depend on contingent combinations of people, timing, and technical feasibility that cannot be reverse-engineered into a simple recipe. Thiel added an explicitly competitive dimension to this view: in many technology categories, being second is functionally equivalent to losing, because the economic rewards can concentrate in a single dominant solution. In such markets, by the time an apparent “trend” becomes legible to a broad audience, it is already too late to treat it as an actionable signal.

Ferguson then shifted from investment practice to Thiel’s more general diagnosis of modernity’s uneven development, associated with Thiel’s argument that the contemporary world displays a “two-speed” pattern: rapid change in digital domains alongside stagnation elsewhere. Thiel grounded the claim in concrete sectors. In transportation, he argued that the long arc of speed improvements that once moved from sailing to rail to automobiles to jet travel had largely plateaued, with the practical experience of travel shaped as much by institutional frictions—security and infrastructure constraints—as by engineering. In energy, he offered a more pointed formulation: if technology is understood as doing more with less, then energy appeared to be moving in the opposite direction, as costs rose and alternatives failed to outcompete incumbent systems. He also acknowledged biomedical progress as a partial exception, while still expressing concern that drug development pipelines and institutional incentives were weakening, suggesting that even in medicine the pace of advances could be less robust than optimistic narratives implied.

Pressed on why innovation seemed concentrated in the “world of bits” rather than the “world of stuff,” Thiel emphasized risk aversion and regulatory accumulation. In his account, heavy regulation in sectors such as nuclear power, aerospace, civil infrastructure, and other “material” domains raised costs, elongated timelines, and narrowed the range of careers and ventures that could plausibly be pursued. He suggested that the resulting labor allocation had cultural and economic consequences: talent flowed into computers and finance in part because these were domains where experimentation was comparatively feasible and returns were attainable, whereas other engineering pathways offered constrained opportunity structures. He noted that alternative explanations also existed—liberal arguments emphasizing underinvestment in science and engineering, and conservative arguments emphasizing the motivating role of sacrifice and geopolitical rivalry—but he treated regulation and risk aversion as the most compelling baseline, while allowing that multiple causal strands could be simultaneously operative.

This diagnosis framed Thiel’s response to a broader challenge: why, if computing is so transformative, do many citizens perceive declining living standards and doubt that the next generation will be better off? Thiel proposed that the tension should not be dismissed as mere misunderstanding; it should instead be treated as an open question about measurement, sectoral distribution, and the relative value of digital progress compared with stagnation in other domains. He offered agriculture as an example: he argued that the productivity surge associated with the mid-20th-century Green Revolution had not been matched in subsequent decades, and he linked the political volatility of food price spikes to this stagnation. The implication was methodological as well as substantive: technological narratives should be integrated with material constraints, price dynamics, and social stability, rather than treated as self-sufficient explanations of political change.

The conversation then moved into political economy and macroeconomic policy. Ferguson framed Thiel’s position as a critique of liberal conventional wisdom that attributes inequality and instability to deregulation, particularly in finance, and that advocates expanded government roles in research, education, and industrial policy. Thiel replied that technology itself should be strongly deregulated and suggested that regulatory barriers in non-digital sectors had indirectly steered talent and capital into domains with different social payoffs. He criticized an approach to crisis management that relies on aggressive monetary expansion and fiscal borrowing while presuming a return to high growth. In his account, Keynesian stabilization works more comfortably in a world of strong technological tailwinds; in a world where progress has slowed, borrowing can become a way of deepening structural fragility rather than bridging a temporary downturn. He illustrated the point by revisiting the 1930s: he argued that the decade’s severe crisis coincided with the emergence of major new industries and technologies, and that these “tailwinds” made certain policy ideas appear more successful than they might otherwise have been. From this angle, debates about austerity versus stimulus were also debates—sometimes implicit—about whether the underlying technological growth regime remained intact.

Audience questions brought the discussion down from macrodiagnosis to applied institutional design. A physician described the fragility of electronic medical record systems and asked how large-scale technology platforms could serve as reliable backups. Thiel refused to discuss specific firms under disclosure constraints but used the question to generalize: he argued that health care, despite extraordinary expense, exhibited surprisingly poor operational technology. He attributed part of this to governance and professional composition: policy debates tend to prioritize distributional questions—coverage and financing—while treating engineering implementation as secondary. He suggested that reform would require technocratic leadership with genuine systems competence, yet observed that administrations of both parties rarely elevate scientists and engineers into decision roles, reflecting a deeper cultural tendency to misclassify engineering as marginal rather than central.

Another question concerned luck in entrepreneurial success. Thiel acknowledged the role of contingency but warned against treating outcomes as lottery tickets. He argued that an emphasis on irreducible luck can become a substitute for disciplined judgment and due diligence. In his own investing practice, he claimed, the weakest decisions were those made with a mindset of blind bets rather than worked-through understanding. He then quantified this with a practical metric—whether valuations increased—and reported that across a large number of investments, somewhat more than half, by that standard, had “worked,” a result he implied was meaningfully different from the distributional intuition suggested by pure chance.

Questions about Europe’s “lost generation” and the post-crisis developed world allowed Thiel to extend his growth thesis. He framed the core issue as restarting growth to avoid a zero-sum politics of redistribution between nations and groups. He also argued that the rhetoric of “developed” versus “developing” countries carries a tacit fatalism: developing countries are imagined as converging by copying, while developed countries are implicitly treated as places where novelty is no longer expected. Thiel urged a conceptual reversal: the task is to “develop the developed world,” meaning to reestablish conditions for genuine technological novelty rather than merely managing mature systems.

Education became a central case study. Thiel argued that, in an environment where broad-based growth is assumed but not realized, societies can generate successive “bubbles” in narratives of progress—misallocated capital chasing socially validated stories. He placed mainstream higher education in this pattern, describing it as overvalued and insulated from return-on-investment scrutiny by social norms that treat such questions as illegitimate. He advised individuals to think explicitly about the purpose of educational pathways rather than treating credential accumulation as an automatic route to future optionality. He described education less as a pure investment or consumption good than as an insurance mechanism—and then criticized the underlying premise, suggesting that a healthier society would ask why the “cracks” are so large that people feel compelled to buy protection from falling through them. He also described schooling as a tournament structure that extends beyond university into professional life, and suggested that institutional leaders rarely speak candidly about the competitive reality because doing so would violate the public moral economy of education.

When asked how to encourage young innovators, Thiel emphasized intrinsic motivation and substantive interest rather than purely status-driven incentives. He pointed to accelerated trajectories—skipping grades, finishing early—as evidence that for highly capable students, the conventional pacing of schooling can reflect institutional inertia rather than pedagogical necessity. He used this to illustrate a broader distinction: if education is treated primarily as competitive positioning, parents will optimize for signaling; if it is treated as learning, parents may accept nonstandard pathways that reduce time spent in credential races. He offered chess as an example of a cognitively demanding activity that can cultivate discipline and strategic thinking, while acknowledging its own risks and idiosyncrasies.

Environmental policy questions elicited a more nuanced position than a simple abolitionist libertarian stance. Thiel recognized that environmental regulation has materially improved certain conditions, such as air pollution, while arguing that regulatory costs are rarely evaluated with commensurate rigor and can foreclose entire categories of technological careers and projects. He proposed nuclear power as a pragmatic, scalable path toward low-carbon electricity, suggesting that climate concerns strengthen the case for an aggressive nuclear buildout. At the same time, he acknowledged that nuclear energy cannot be wholly deregulated due to waste and proliferation risks, and he emphasized that regulatory design—timelines, siting, veto points—can be decisive. In his view, delays stretching to decades, or the capacity of narrow political actors to block waste disposal solutions, can functionally prevent the technology from being deployed at scale.

A question about “open intelligence” and data-driven insight led Thiel to outline a thesis about the “big data” problem: modern institutions possess vast information but lack usable mechanisms for synthesis, interpretation, and action. He described opportunities for both enterprise and consumer applications, ranging from organizational situational awareness to personalized recommendation systems that anticipate needs users may not yet articulate. In explaining Palantir, he characterized its origins as an attempt to transfer fraud-detection techniques into national security and complex organizational contexts, where connecting dispersed data points might prevent terrorist plots or other forms of organized crime. The moderator raised a civil-liberties concern: tools that make the state more informationally capable could, in principle, enable an omniscient surveillance apparatus. Thiel responded with a consequentialist comparison: absent effective technological tools, a major attack could produce a far more sweeping and blunt political response, with greater erosions of liberty than a targeted analytic capability designed to prevent the event in the first place. In this calculus, the risk of empowering institutions had to be weighed against the risk of crisis-driven overreaction.

Later questions returned to the problem of government-led innovation and why it appears less capable of path-breaking projects than in wartime eras. Thiel suggested that contemporary governance often treats technology as something that will “work on its own,” so it does not require deep substantive engagement. He criticized “portfolio theory” approaches to public investment in clean technology that disperse funds across many initiatives without serious internal evaluation of engineering feasibility. Using the example of Solyndra, he argued that political defenses and critiques alike often avoid the central technical question—what actually works and why—because it is more difficult than procedural or ethical arguments. In his view, both parties tend to be more comfortable litigating processes than confronting engineering realities, which diminishes the state’s capacity to sponsor breakthroughs.

A final audience exchange pushed the time horizon to genetic engineering and the prospect of radically amplifying intelligence over centuries. Thiel responded skeptically, arguing that most of human history has been static rather than relentlessly progressive, and that the modern era’s rapid technological acceleration is historically exceptional rather than a guaranteed baseline. For that reason, he cautioned against treating exponential progress as an automatic default that will be replaced by an even faster regime if current engines slow. The practical implication was consistent with his earlier claims: if technological deceleration is real, it should be treated as a central policy and civilizational problem, not as a temporary anomaly that will resolve itself.

The event concluded with a question about entrepreneurship and the decision to sell a company, prompted by PayPal’s sale to eBay. Thiel described the sale as shaped by competitive threats from entrenched platforms and payment networks, and he broadened the answer into an institutional theory of innovation. He argued that companies have a “founding period” during which innovation is concentrated, and that transitions to bureaucratic management can truncate this period and reduce inventive capacity. He used Apple’s mid-1980s governance decision to remove Steve Jobs as an illustrative case of premature closure of a founding cycle, and he argued that founder-led firms are more likely to sustain innovation. From that perspective, preserving founder control is not merely a matter of personal authority; it is a structural condition for extending the window in which an organization behaves like an invention machine rather than a stable administrative apparatus.

Across its topics, the conversation presented a coherent diagnostic posture: the modern world displays highly visible progress in computational domains while many material sectors exhibit slowdown; politics and public administration often treat engineering substance as secondary; and societies that assume automatic growth can misallocate capital into socially legitimized narratives—whether in financial markets, housing, or credential systems—rather than confront the harder work of rebuilding conditions for vertical technological advance. The exchange did not resolve these claims so much as stage them as testable hypotheses about institutions, incentives, and the distribution of technological dynamism, with the repeated insistence that restarting growth requires an explicit confrontation with what has become difficult to do, and why.