Peter Thiel's Principles
A Benchmark for AI's Creative Pursuit of Intellectual Truth in Philosophy
Summary
Since the 1970s, digital technology has provided vast access to information, yet transforming this into impactful ideas remains a distinct human strength. Thinkers like Yuval Noah Harari and Peter Thiel exemplify this ability. This essay proposes that Thiel’s contrarian, multidisciplinary principles offer a benchmark for measuring AI’s creativity as it approaches artificial general intelligence (AGI). Enduring human achievements in philosophy rely on objective truth, history, empathy, and intuition. For AI to shape humanity’s future, it must meet this same standard of philosophical creativity.
1. Introduction
Artificial intelligence (AI) is advancing rapidly. Models like xAI’s Grok-4, Google DeepThink, and OpenAI’s ChatGPT-5 have achieved remarkable milestones. For instance, Grok-4 leads the ARC-AGI-3 Leaderboard with approximately 16% accuracy, a significant lead over competitors, who remain below 10% (Shah. A, 2025) [1]. It also scored a perfect 100% on the 2025 AIME competition [2], a benchmark for high school math talent. Meanwhile, in experimental settings, ChatGPT-5 and DeepThink solved five of six problems at the 2025 International Mathematical Olympiad (IMO) [3], earning Gold Medals, surpassing Google DeepMind’s 2024 Silver Medal (Bischoff, M. 2024) [4]. These AI systems used natural language reasoning and proof-writing without human assistance (Roth, 2025 [5]; Roth, 2025 [6].
In contrast, most widely available free-tier large language models (LLMs) (Zhao et al., 2023) [7] typically solve only the first two IMO problems. Notably, roughly 5% of human IMO participants solved the hardest sixth problem, highlighting the complex areas where human intelligence still holds an advantage.
This rapid progress in AI’s mathematical skills illustrates a Thielian insight (Thiel & Cowen, 2024 @50:30) [8]: as AI masters well-structured problem-solving, specialists in these domains may face faster displacement than those engaged in the more complex, less quantifiable realms of human experience and creative thought.
2. The Challenge of Defining AGI Success
Solving mathematical and scientific problems has become routine for top-tier AI, yet these achievements don’t encompass the full spectrum of human intelligence. A deeper question arises: when will AI truly surpass humanity’s intellectual best? Nick Bostrom defines superintelligence as an intellect that "greatly exceeds the cognitive performance of humans in virtually all domains of interest" (Bostrom, 2014) [9]. AGI, in my view, will only be reached when AI models produce original, lasting philosophical ideas, those that have historically shaped civilizations and humanity’s self-understanding. Though current large language models (LLM) excel at synthesizing and expressing information, qualities like empathy, intuition, and ethics still form the basis for creating profound and enduring thought.
3. Philosophy as the Apex of Human Achievement
The Doctor of Philosophy (Ph.D.) remains academia’s summit across most disciplines, except for medicine. Yet, the title itself is no guarantee of foundational impact. Figures like Thomas Aquinas (Summa Theologica, 1225–1274) [10], who developed profound philosophical and theological systems; René Descartes (1596–1650), the father of modern philosophy and rationalism (Discourse on Method, 1637) [11]; and John Locke (1632–1704), a key architect of modern liberal thought (Two Treatises of Government, 1689) [12] profoundly shaped theology, rationalism, and liberal thought. Their contributions illustrate that lasting philosophical influence does not require formal credentials.
This emphasis on “philosophy” as the apex of intellectual achievement reveals a belief that philosophical insight surpasses empirical or technical mastery. History’s greatest civilizations have been shaped by foundational philosophical frameworks, from Greek, Roman, Eastern (including Indian and East Asian), Middle Eastern, and African thinkers, through the Renaissance and Enlightenment, to the Industrial Revolution and today’s AI Age. These ideas are often as hard-won and consequential as groundbreaking scientific discoveries (Floridi, 2019) [13]. Unlike empirical proofs, they are forged through conceptual rigor and lasting relevance.
4. Measuring AI Creativity: The Peter Thiel Benchmark
How do we meaningfully measure philosophical creativity in AI? Peter Thiel’s wide-ranging, contrarian worldview provides a rich benchmark. He is renowned for synthesizing insights from philosophy, history, religion, economics, and technology, boldly tackling society’s toughest questions and embracing intellectual risk and complexity (Chafkin, 2021) [14]. His approach actively challenges consensus, drawing inspiration from thinkers like Leo Strauss's Natural Right and History (Strauss, 1953) [15], Carl Schmitt's Political Theology (Schmitt, 1985) [16], and René Girard's Violence and the Sacred (Girard, 1977) [17], while warning against the dangers of groupthink and stagnation. Testing AI against Thiel’s multidimensional framework helps us assess whether machines can genuinely transcend knowledge synthesis to produce original, impactful philosophical thought.
5. Seven Core Principles from Thiel’s Thought
Drawing on Peter Thiel’s extensive body of his own works (including Zero to One (Thiel, 2014) [18]) and public statements (startuparchive, 2024) [19] (e.g., Thiel & Robinson, 2019-2024 in Uncommon Knowledge [20]-[25], (Atlas Society, 2023) [26])), supplemented by insights from biographies such as The Contrarian (Chafkin, 2021) [14], this section presents seven principles that define his worldview, each infused with unconventional insights igniting debate and creativity. Some are more directly stated (notably 5.2 through 5.5 and 5.7); others require deeper interpretation (especially 5.1 and 5.6).
5.1 Stagnation vs. Progress
Thiel stresses that real societal and technological progress requires creating fundamentally new things, a leap from “zero to one,” rather than just improving what already exists. He contrasts those leaps with the repeatable process of "one to N" (Thiel, 2013; @3:00) [27]. For AI, this means novel algorithms, not copying rivals’ models.
True Progress Needs Monopolies: Creative monopolies, not endless competition or exploitative rent-seeking, drive true progress. Competition often leads to zero-sum struggles that waste value and cause stagnation. Thiel is among the first to note that technological stagnation, often contrasted as atoms versus bits, began in the 1970s outside of computing and communications (Thiel, 2011 [28]; Thiel & Andressen, 2013; @7:20 [29]; Thiel & Robinson, 2019 [18]). His statement, “We wanted flying cars, instead we got 140 characters” (Packer, 2011) [30], reflects his concern that society has pulled back from bold innovation. He often highlights this trend by contrasting ambitious endeavors like SpaceX’s reusable rockets with the more limited advancements seen in digital platforms like Twitter (Fowler, 2017; @13:30) [31].
Illusions of Progress: Thiel warns against mistaking horizontal scaling (simply expanding or copying existing technologies) for real innovation. He reflects on J.J. Servan-Schreiber’s 1968 book The American Challenge (Servan-Schreiber, 1968) [32], which urged Europe to adopt American dynamism. Ironically, Thiel says America later failed to keep that innovative edge.
Academic Drift: He also notes that core fields like physics and chemistry are universally recognized as true sciences and have produced genuine breakthroughs. By contrast, newer disciplines, such as social science, political science, or environmental science, often append the term “science” to their names in an attempt to gain similar prestige, even when their methods or innovations may lack the depth or rigor of the natural sciences (Thiel, 2021; @10:20) [26]. To Thiel, this shows society’s retreat from meaningful progress. Indeed, Thiel’s early critique, articulated nearly thirty years ago in The Diversity Myth (Sacks & Thiel, 1998) [33], signaled his concerns about how ideological conformity, even under the guise of progress, can stifle genuine intellectual and institutional innovation. This critique served as a harbinger of the broader stagnation he now observes.
5.2 Visionary Leadership
Thiel believes major breakthroughs come from risk-taking visionaries, or “definite optimists,” who push forward with strong, sometimes “cult-like” conviction. Though this kind of leadership fuels innovation, he acknowledges the dangers when unchecked power emerges. His focus on definite optimism echoes Calvinist ideas about a predetermined, purposeful path guided by strong calling.
5.3 Contrarian Discovery
For Thiel, innovation means seeking “secrets” or unconventional truths ignored or rejected by the mainstream. However, he emphasizes that unconventional contrarianism by itself is not sufficient, but must be grounded in intellectual truth with merit (Fowler, 2017; @5:20 [31]; Morgan, 2023; @7:50 [34]). This requires intellectual bravery and social risk-taking, as detailed in his book Zero to One [18]. In later work, he connects discovery to humans’ deep need for shared meaning, often linked to religion.
He warns against two extremes: one is excessive skepticism or scientism, which rejects anything that cannot be empirically proven; the other is rigid dogmatism, which refuses to consider new or challenging ideas (Thiel, 2021; @11:10) [24]. True progress, he argues, depends on navigating between these pitfalls. For AI, this means questioning data-driven biases while also being open to genuinely novel insights.
5.4 Imitation and Rivalry
Human behavior is strongly influenced by mimetic desire, the tendency to want what others want, which explains how trends rapidly spread through fashion, social media, and other social networks. This imitation can escalate into rivalry and destructive competition in social and economic arenas (Girard, 1977 [17]; Thiel, 2014 [18]).
Building on Girard’s insights, Thiel highlights how such mimetic rivalry fuels much of modern business competition, such as between Uber and Lyft, as well as broader social conflicts. These dynamics often result in wasted effort, lost value, and scapegoating as a form of social release. Recognizing these dynamics shows why truly new innovation, the leap from “zero to one,” is crucial for avoiding the pitfalls of mere imitation and fostering genuine progress.
5.5 Limits of Modern Democracy
Thiel critiques modern democracies for often producing bureaucratic stagnation that blocks true freedom and market efficiency (Chafkin, 2021) [14]. He warns that unchecked democracy, especially when focused on redistributing wealth rather than creating value, can backfire.
Instead, he explores alternative governance models, such as market-based or technology-managed systems, which he believes better support progress and personal liberty. Thiel’s critique, inspired by Strauss, suggests that scientific rationalism, dominant since the French Revolution, displaced classical political wisdom, harming authentic leadership (Strauss, 1953) [15].
5.6 Dangers of Centralization
Thiel is deeply worried that too much centralization, especially on a global scale, threatens liberty and innovation. Concentrated power leads to inefficiency, bureaucracy, and systemic risks that could be existential.
To protect freedom and foster progress, he calls for bold, decisive, and sometimes unconventional leadership (Chafkin, 2021) [14]. Thiel consistently favors “vertical” progress (making truly new things) over “horizontal” expansion (just growing what already exists), warning the latter often hides stagnation.
Religion and Leadership amid Centralization and Crisis: Thiel’s thinking is influenced by controversial theorists like Carl Schmitt whose works, including The Concept of the Political (Schmitt, 2005) [35] and especially his concept of Political Theology (Schmitt, 1985) [16], highlight the need for strong, sovereign decisions during crises. This supports Thiel’s call for decisive leadership in exceptional times.
His worldview also has Christian elements, warning against a unifying, technocratic “Antichrist” force that suppresses individuality, freedom, and true innovation (Thiel & Douthat, 2025, @45:50) [36]. He sees AI, due to its disruptive impact, as a double-edged sword, offering great progress but also introducing the risk of authoritarian control. His 2018 phrase “If crypto is libertarian, AI is communist” highlights this tension (Thiel & Cowen, 2024; @54:20) [8]. Even as Thiel initially supported cryptocurrency for decentralization, he observes growing centralization in crypto exchanges, regulations, and mining, reinforcing concerns about concentrated power.
Thiel also refers to the biblical Katechon, a restraining force that holds back chaos, suggesting principled leadership is needed to avert social and political collapse (Thiel & Cowen, 2024; @37:10) [8].
Modern State Control from Historical Perspective: Applying a Straussian lens, Thiel argues that since the French Revolution, there’s been a shift in who gains power. A common belief at that time was that mathematical ability, unlike verbal skill, was evenly distributed and more “meritocratic.” As a result, modern states began preferring “math people”: scientists, engineers, and technocrats. However, Thiel warns these “math people,” though often brilliant in special fields, are often less attuned to the full complexities of humanity. This can make them easier for centralized authorities and bureaucracies to manage or exploit, like the Soviet Union’s promotion of chess masters and mathematicians as models of state excellence (Thiel & Cowen, 2024; @51:00) [8]. Thiel views this technical expertise focus as part of a cold, controlling form of government valuing order and predictability. He believes this shift has made society less vibrant and more prone to decline, rather than renewal.
5.7 Overcoming Paralysis
Thiel’s framework contrasts determinate vs. indeterminate optimism and pessimism (Thiel, 2013; @12:00) [27]. Both unfocused or indefinite optimism (expecting improvement without real planning) and deep pessimism (believing nothing can be done) lead to societal paralysis (Sulak, 2024) [37]. Blind hope and hopeless despair both stifle action. True progress requires a clear, focused vision of the future combined with determined effort to make it real. AI needs focused design to avoid aimless outputs.
5.8 Sources
This summary is mainly based on Thiel’s Zero to One [18] (especially on visionary leadership, contrarian discovery, and imitation and rivalry, which cover 5.2-5.4 mainly), along with his interview quotes compiled by Perell (2019) [38], his biography (Chafkin, 2021) [14], Stephen Diehl’s critiques of Thiel’s worldview (Diehl, 2021) [39], and numerous public speeches and interviews on political theology and philosophy of science. Despite prioritizing Thiel’s earliest possible articulations, I have also included references to later instances where he directly reiterated these ideas across his prolific speeches and interviews.
6. Why Philosophy Matters for AGI
Measuring scientific or mathematical skill is relatively straightforward. Evaluating philosophical or creative impact is far more subtle and subjective, often judged differently across generations. Benchmarks like Peter Thiel or Yuval Harari suggest a way to test if AI can originate ideas that truly shape society, rather than just recombine existing knowledge. Enduring ideas typically challenge consensus and reshape understanding (Boden, 2004 [40]; Colton et al., 2009 [41]). Guiding AI toward “truthfulness” in complex, subjective areas therefore sets a high philosophical goal. Meeting such objective requires self-verification, self-improvement (or autonomous innovation), and the ability to refine its models of deep, contested truths over time (Christian, 2020 [42]; Liu et al., 2025, [43]).
For many, AI’s highest value is helping people see the world anew, enlightening, surprising, and inspiring, as explored in my earlier piece “Is AI Chatbot My Boswell?” (Luh, 2025) [44]. If AI achieves this, it moves closer to the threshold of true AGI.
7. The Difficulty of Measuring AI Originality
As the saying goes, “beauty is in the eye of the beholder.” So is creativity. Humans combine knowledge into genuinely original insights, whereas today’s most advanced AI often rearranges and curates, rather than originates. We still await the moment when AI produces a philosophical framework as novel and influential as those throughout history.
A key difference lies in empathy and intuition. Beyond understanding suffering, hopes, and morality, philosophical insight also springs from intuition’s flashes. These intuitive moments let humans recognize patterns, sense deeper significance, and form well-rounded judgments without conscious reasoning, complementing empathy’s role in connecting with others’ experiences. Together, these traits enable humans to generate nuanced, original ideas that resonate both intellectually and emotionally.
Though current large language models (LLM) can mimic empathetic responses, they lack genuine moral and emotional understanding, an important limitation documented in recent research on affective computing and AI ethics (Bender et al., 2021 [45]; Weidinger et al., 2021 [46]; Picard, 1997 [47]). Replicating such intuitive insight remains a significant challenge for AI (Russell, 2019) [48]. Ultimately, this blend of intellectual insight and emotional depth may prove to be the defining characteristic of genuine philosophical creativity in AGI.
8. Discussion
Our contemporary thinkers like Harari or Thiel, of course, haven't yet reached the stature of philosophical giants like Aquinas, Descartes, or Locke due to their short history. Their originality, however, lies in their remarkable ability to sift insights that are in plain sight. They possess such a profound grasp of historical significance that they can fluidly weave diverse historical contexts into new ideas, illuminating present challenges. Yet, only a select few individuals can produce such influential, original ideas. Among these, Thiel's work is particularly noteworthy for its consistent focus on intellectual truth as he constructs his framework for human advancement and technological progress. This rare synthesis is precisely what I hope AGI could one day achieve, encompassing both the pursuit of truth and the embodiment of benevolence. Hence, I have selected Thiel’s principles as models for AI to emulate and learn in the pursuit of genuine philosophical creativity.
Looking ahead: Should AGI begin generating profound new philosophical ideas at a pace beyond all historical precedent, potentially producing transformative insights daily or even hourly, humanity will face a unique challenge. Absorbing, interpreting, and integrating this stream of breakthrough thought may become as vital as any technological advance. This process requires both individual and collective adaptation to a world continuously reshaped by emergent philosophical paradigms. Open questions then arise: How might human minds tolerate or be overwhelmed by such relentless intellectual stimulation? Or, worse yet, how might AGIs cope with each other amid a proliferation of new AGIs?
9. Conclusion
AI’s progress in mathematics, science, and language is accelerating rapidly. Yet true artificial general intelligence (AGI) will ultimately be judged by its meaningful contributions to the creative and philosophical essence of humanity. Benchmarks inspired by Peter Thiel’s principles offer a framework for evaluating whether AI can achieve this level of depth. Until then, the art of creating lasting philosophical insight, rooted deeply in empathy, intuition, and history, remains uniquely human. History provides the essential context that grounds our empathy and intuition, so we can learn from the past and shape the future. This capacity is vital not only for advancing AI, but for the ongoing evolution of society itself. Will AI ever truly grasp the dynamic interplay of empathy, intuition, and historical understanding that defines our humanity, or will these remain the final bastions of human creativity?
Hashtags:
#ArtificialIntelligence, #AICreativity, #AGI, #PeterThiel, #Philosophy, #Innovation, #TruthInAI, #ContrarianThinking, #LLM, #FutureOfAI, #AIResearch, #AIReasoning
Acknowledgments:
Thanks to a reader’s feedback on “Invisible Hand of AI Resonance” [49], inspiring me to explore Thiel’s ideas. Thanks are also to Google Gemini 2.5 Flash, xAI Grok 3, Perplexity AI, and Qwen 2.5-Max for their assistance in checking accuracy and streamlining my writing for enhanced readability. The essay’s cover image of Peter Thiel is a cropped still from his Zero to One [18] interview on Socrates in the City’s YouTube channel [50].
References:
[1] Shah, A. (2025). Grok 4 AI Models Set New Benchmark Records, theaisignal.com, July 2025.
[2] Mathematical Association of America, AIME Competition, competitionsciences.org.
[3] International Mathematical Olympiad, IMO, imo-official.org.
[4] Bischoff, M. (2024). AI Reaches Silver-Medal Level at This Year’s Math Olympiad, scientificamerican.com.
[5] Roth, W. (2025). OpenAI Just Solved Math. (2025). YouTube @2:40, July 2025.
[6] Roth, W. (2025). Google Takes Gold. (2025). YouTube @3:20, July 2025.
[7] Zhao, W. et al., (2023). A Survey of Large Language Models, arXiv.org; May 2023; see also Large Language Models, en.wikipedia.org; Jun 2025.
[8] Thiel, P. & Cowen, T. (2024). Peter Thiel on Political Theology | Conversations with Tyler, YouTube @12:45 (scapegoating), @37:10 (Katechon), @50:30 (worse for math people), @51:00 (French Rev), @52:00 (math people), @54:20 (libertarian-crypto), Mercatus Center; see also transcript, conversationswithtyler.com.
[9] Bostrom, N. (2014). Superintelligence: Paths, Dangers, Strategies, Oxford University Press, book.google.com, 2014; see also wikipedia, Superintelligence: Paths, Dangers, Strategies, wikipedia.org, July 2025.
[10] Aquinas, T. (1265–1274). Summa Theologica, archive.org.
[11] Descartes, R. (1637). Discourse on the Method, archive.org.
[12] Locke, J. (1689). Two Treatises of Government, yorku.ca.
[13] Floridi, L. (2019). The Ethics of Artificial Intelligence, Oxford University Press.
[14] Chafkin, M. (2021). The Contrarian: Peter Thiel and Silicon Valley's Pursuit of Power, Penguin Press, google.com.
[15] Strauss, L. (1953). Natural Right and History, University of Chicago Press.
[16] Schmitt, C. (1985). Political Theology: Four Chapters on the Concept of Sovereignty, MIT Press.
[17] Girard, R. (1977). Violence and the Sacred, Johns Hopkins University Press.
[18] Thiel, P., & Masters, B. (2014). Zero to One: Notes on Startups, or How to Build the Future, Crown Business.
[19] startuparchive. (2024). The essays, talks, and interviews of Peter Thiel, startuparchive.org.
[20] Thiel P. & Robinson, P. (2019). Peter Thiel on “The Straussian Moment,” Hoover Institution; see also YouTube: @8:05 (Woodstock 1969).
[21] Thiel, P. & Robinson, P. (2020). The World According To Thiel, Hoover Institution.
[22] Thiel, P. & Robinson, P. (2021). Make Ticker Tape Parades Great Again: A Conversation With Peter Thiel, Hoover Institution.
[23] Thiel, P. & Robinson, P. (2022). Peter Thiel, Leader Of The Rebel Alliance, Hoover Institution; see also Robinson, P. (2022). Peter Thiel, Leader of the Rebel Alliance, nationalreview.com.
[24] Thiel, P. & Robinson, P. (2024). Apocalypse Now? Peter Thiel on Ancient Prophecies and Modern Tech, Hoover Institution.
[25] Thiel, P. & Robinson, P. (2024). Part II: Apocalypse Now? Peter Thiel on Ancient Prophecies and Modern Tech, Hoover Institution.
[26] Thiel, P. (2014). The 2021 Atlas Society Gala: Peter Thiel's Speech; YouTube @10:20 (science?), @11:00 (dogma-skepticism), Atlas Society.
[27] Thiel, P. (2013). Peter Thiel: You Are Not a Lottery Ticket | Interactive 2013 | SXSW, YouTube @3:00 (1-N), @12:00 (determinate optimism), @29:40 (polling), @35:40 (actuary), SXSW, Oct 2013.
[28] Thiel, P. (2011). The End of the Future, nationalreview.com.
[29] Thiel, P. & Anressen, M. (2013). Copy of In Tech We Trust? A Debate with Peter Thiel and Marc Andreessen, YouTube @7:20, Milken Institute, May 2013.
[30] Packer, G. (2011). No Death, No Taxes, newyorker.com.
[31] Fowler, M. (2017). Peter Thiel at Launch 2015, YouTube @13:30 (innovation-bits).
[32] Servan-Schreiber, J.-J. (1968). The American Challenge, Atheneum.
[33] Sacks, D. & Thiel, P. (1998). The Diversity Myth, independent.org.
[34] Morgan, C. (2023). The Untold Story of Peter Thiel, YouTube @5:50 (contrarian).
[35] Schmitt, C. (2005). The Concept of the Political, University of Chicago Press.
[36] Thiel, P. & Douthat, R. (2025). A.I., Mars and Immortality: Are We Dreaming Big Enough? | Interesting Times with Ross Douthat, YouTube @45:35 (AntiChrist), June 2025.
[37] Sulak, J. (2024). Definite vs. indefinite Thinking: Notes from Zero to One by Peter Thiel, boxkitemachine.net.
[38] Perell, D. (2019). Peter Thiel: Uncommon Knowledge, perell.com.
[39] Diehl, S. (2021). Deconstructing Thiel, Deconstructing the Worldview of Peter Thiel, stephendiehl.com.
[40] Boden, M. A. (2004). The Creative Mind: Myths and Mechanisms, Routledge.
[41] Colton, S., et al. (2009). Computational Creativity: Coming of Age, AI Magazine, 30(3), 11–14.
[42] Christian, B. (2020). The Alignment Problem: Machine Learning and Human Values, W. W. Norton.
[43] Liu, Y. et al. (2025). AlphaGo Moment for Model Architecture Discovery, arXiv.org, July 2025.
[44] Luh, P. (2025). Is AI Chatbot My Boswell?: Testing for Chatbots Becoming Indispensable, a Boswell Test, substack.com, Feb 2023.
[45] Bender, E. M., et al. (2021). On the Dangers of Stochastic Parrots, FAccT ’21.
[46] Weidinger, L., et al. (2021). Ethical and social risks of harm from Language Models, arXiv.org.
[47] Picard, R. W. (1997). Affective Computing, MIT Press.
[48] Russell, S. (2019). Human Compatible: Artificial Intelligence and the Problem of Control, Viking.
[49] Luh, P. (2025). Invisible Hand of AI Resonance: Human-AI Interaction, from Pendulum to Propaganda, substack.com, July 2025.
[50] Thiel, P. & Metaxas, E. (2020). Peter Thiel: Zero to One, Socrates in the City, Feb 2020.