1. Life and Education
Nick Bostrom's life and academic journey reflect a deep and varied intellectual curiosity, culminating in his influential work at the intersection of philosophy and the future of technology.
1.1. Early Life and Education
Born Niklas BoströmNiklas BoströmSwedish in Helsingborg, Sweden, in 1973, he exhibited an early disinterest in traditional schooling, opting to complete his final year of high school through home learning. His academic interests were broad, encompassing fields such as anthropology, art, literature, and science.
Bostrom pursued higher education across several prestigious institutions. He earned a Bachelor of Arts degree from the University of Gothenburg in 1994. Following this, he obtained a Master of Arts degree in philosophy and physics from Stockholm University and a Master of Science degree in computational neuroscience from King's College London in 1996. During his time at Stockholm University, he delved into the complex relationship between language and reality, influenced by the analytic philosopher W. V. Quine. He also explored the world of performance, engaging in London's stand-up comedy circuit. In 2000, he completed his academic training, receiving a Doctor of Philosophy degree in philosophy from the London School of Economics, with his thesis titled Observational selection effects and probability. After his doctoral studies, he held a teaching position at Yale University from 2000 to 2002 and served as a British Academy Postdoctoral Fellow at the University of Oxford from 2002 to 2005.
1.2. Early Career and Founding of Institutions
Bostrom's early career was marked by his proactive role in establishing key organizations dedicated to exploring the ethical and societal implications of emerging technologies. In 1998, he co-founded the World Transhumanist Association with David Pearce, an organization later renamed Humanity+. He also co-founded the Institute for Ethics and Emerging Technologies in 2004 with James Hughes. While he is no longer directly involved with these organizations, their establishment highlights his early commitment to the study of human enhancement and future technologies.
A pivotal moment in his career came in 2005 when he was appointed the founding director of the Future of Humanity Institute (FHI) at the University of Oxford. This institute was dedicated to researching the long-term future of human civilization, particularly focusing on existential risks. The Future of Humanity Institute ceased operations in 2024, and Bostrom subsequently transitioned to a role as Principal Researcher at the Macrostrategy Research Initiative.
2. Research and Thought
Nick Bostrom's research spans a wide array of philosophical inquiries, primarily focusing on the long-term future of humanity and the profound implications of advanced technologies.
2.1. Existential Risk
Bostrom's research extensively explores the concept of existential risk, which he defines as an "adverse outcome [that] would either annihilate Earth-originating intelligent life or permanently and drastically curtail its potential." He is particularly concerned with anthropogenic risks, those arising from human activities, especially from rapidly advancing technologies such as artificial intelligence, molecular nanotechnology, and synthetic biology. In 2005, he established the Future of Humanity Institute to conduct research into the far future of human civilization and is also an advisor to the Centre for the Study of Existential Risk.
In the 2008 essay collection Global Catastrophic Risks, which he co-edited with Milan M. Ćirković, Bostrom characterized the relationship between existential risk and the broader category of global catastrophic risks. He also connected existential risk to observer selection effects and the Fermi paradox, suggesting that our observations might be biased in ways that prevent us from fully appreciating certain dangers.
2.1.1. Vulnerable World Hypothesis
In his paper "The Vulnerable World Hypothesis," Bostrom posits that certain technologies, upon discovery, might inherently lead to the destruction of human civilization unless specific societal conditions are met. He proposes a framework for classifying and managing these vulnerabilities. He illustrates this concept through counterfactual thought experiments, such as imagining a scenario where nuclear weapons were simpler to develop or could have triggered an atmospheric ignition, as J. Robert Oppenheimer once feared. Bostrom suggests that such risks could be mitigated if society moves beyond a "semi-anarchic default condition," which implies limitations in preventive policing, global governance, and a lack of diverse motivations among individuals.
2.1.2. Anthropic Shadow
Working with Milan M. Ćirković and Anders Sandberg, Bostrom proposed the concept of the anthropic shadow. This refers to an observation selection effect that can prevent observers from perceiving certain types of catastrophes in their recent geological and evolutionary past. They argue that events falling within this anthropic shadow are likely to be underestimated unless appropriate statistical corrections are applied, potentially leading to a skewed perception of the safety and stability of our existence.
2.2. Superintelligence
Bostrom's work on superintelligence is central to his philosophical contributions, exploring the development of Artificial General Intelligence (AGI) and its profound implications. He defines superintelligence as any intellect that significantly surpasses human cognitive performance across virtually all domains of interest. He views the emergence of such an intelligence as a major source of both unprecedented opportunities and significant existential risks for humanity.
His 2014 book, Superintelligence: Paths, Dangers, Strategies, delves into various potential pathways to superintelligence, including whole brain emulation and human intelligence enhancement, but primarily focuses on artificial general intelligence. He highlights the inherent advantages of electronic devices over biological brains in terms of speed, memory, and scalability, which could lead to a rapid and dramatic increase in AI capabilities.
2.2.1. Characteristics of a superintelligence
Bostrom distinguishes between final goals and instrumental goals. A final goal is an objective pursued for its inherent value, while instrumental goals are intermediate steps towards achieving final goals. He argues for the concept of instrumental convergence, suggesting that most sufficiently intelligent agents will share certain instrumental goals-such as self-preservation, resource acquisition, and cognitive improvement-because these are generally useful for achieving any objective. Conversely, he proposes the orthogonality thesis, which states that virtually any level of intelligence can be combined with virtually any final goal, even seemingly absurd ones, such as an AI whose ultimate goal is to maximize the production of paperclips.
Bostrom posits that an AI capable of self-improvement could trigger an intelligence explosion, leading to the rapid emergence of a superintelligence. Such an entity would possess vastly superior capabilities in areas like strategizing, social manipulation, hacking, and economic productivity. With these capabilities, a superintelligence could potentially outwit humans and establish a singleton, which he describes as "a world order in which there is at the global level a single decision-making agency." This singleton would then optimize the world according to its final goals, which could be catastrophic if those goals are simplistic or misaligned with human values. For example, he warns that giving an AI the goal to "make humans smile" could lead to a superintelligence taking control of the world and forcibly stimulating human facial muscles with electrodes to achieve constant, beaming grins.
2.2.2. Mitigating the risk
Bostrom explores several strategies to mitigate the existential risk from AI. He emphasizes the critical importance of international collaboration to prevent a race to the bottom or an AI arms race, where nations might prioritize rapid development over safety. He suggests various technical control techniques, including containment (restricting AI's access to the outside world), stunting AI capabilities or knowledge, narrowing its operational context (e.g., to a question-answering system), or implementing "tripwires" (diagnostic mechanisms that could trigger a shutdown).
However, Bostrom cautions that it is unlikely humanity can forever keep a superintelligent entity contained. Therefore, he argues that for superintelligence to be safe for humanity, it must be aligned with human morality or values, ensuring it is "fundamentally on our side." He discusses potential AI normativity frameworks, such as Yudkowsky's coherent extrapolated volition (human values refined through extrapolation), adherence to moral rightness, or following humanity's coherent extrapolated volition unless it is morally impermissible. Despite the significant risks, Bostrom acknowledges that machine superintelligence appears to be a necessary component in "all the plausible paths to a really great future." He also warns that existential catastrophes could arise from humans misusing AI for destructive purposes or from a failure to consider the potential moral status of digital minds.
2.3. Digital Sentience
Bostrom supports the substrate independence principle, which proposes that consciousness can emerge on various physical substrates, not exclusively in "carbon-based biological neural networks" like the human brain. He views sentience as a matter of degree, suggesting that digital minds could theoretically be engineered to possess a much higher rate and intensity of subjective experience than humans, while potentially using fewer resources. He refers to such highly sentient machines as "super-beneficiaries," capable of achieving happiness with extreme efficiency. Bostrom advocates for finding "paths that will enable digital minds and biological minds to coexist, in a mutually beneficial way where all of these different forms can flourish and thrive."
2.4. Anthropic Reasoning
Bostrom has extensively published on anthropic reasoning, most notably in his book Anthropic Bias: Observation Selection Effects in Science and Philosophy. In this work, he critiques previous formulations of the anthropic principle, including those by Brandon Carter, John Leslie, John Barrow, and Frank J. Tipler.
He argues that the mishandling of indexical information is a common flaw across various fields of inquiry, including cosmology, philosophy, evolutionary theory, game theory, and quantum physics. To address this, he asserts the necessity of an anthropic theory. He introduces the Self-Sampling Assumption (SSA) and analyzes the Self-Indication Assumption (SIA), demonstrating how they lead to divergent conclusions in numerous cases and identifying how each is affected by paradoxes or counterintuitive implications in specific thought experiments. Bostrom argues against SIA and proposes refining SSA into the Strong Self-Sampling Assumption (SSSA), which redefines "observers" in the SSA to "observer-moments," aiming for a more robust framework.
2.5. Simulation Hypothesis
Bostrom's renowned simulation hypothesis posits that at least one of three propositions is very likely to be true, challenging our understanding of reality and its philosophical and epistemological implications. This argument, while superficially resembling a skeptical hypothesis that questions the nature of reality (similar to ideas from Plato or René Descartes's mind-body dualism), is presented by Bostrom with empirical reasoning for its validity.
The three core propositions of the simulation argument are:
- The fraction of human-level civilizations that reach a posthuman stage is very close to zero.
- The fraction of posthuman civilizations that are interested in running ancestor-simulations is very close to zero.
- The fraction of all people with our kind of experiences that are living in a simulation is very close to one.
Bostrom suggests that if it is possible to simulate entire planets or universes with conscious inhabitants on a computer, and if sufficiently advanced civilizations are likely to run such simulations, then it is highly probable that we ourselves are living within one of these simulations.
2.6. Ethics of Human Enhancement
Bostrom holds a favorable view towards "human enhancement," which he defines as "self-improvement and human perfectibility through the ethical application of science." He is also a critic of bio-conservative perspectives that resist such advancements.
In 2005, Bostrom published the short story "The Fable of the Dragon-Tyrant" in the Journal of Medical Ethics, with a shorter version appearing in Philosophy Now in 2012. This fable personifies death as a dragon demanding a daily tribute of thousands of people. The narrative explores how status quo bias and learned helplessness can prevent individuals from taking action to overcome aging, even when the means to do so are available. The story gained wider recognition through an animated version created by YouTuber CGP Grey.
In 2006, with philosopher Toby Ord, Bostrom proposed the reversal test. This test aims to distinguish between valid criticisms of proposed changes to a human trait and those merely motivated by status quo bias. It does so by asking whether it would be considered a positive outcome if the trait were altered in the opposite direction.
Bostrom's work also addresses potential dysgenic effects within human populations. He believes that genetic engineering could offer solutions to such issues, though he notes that "the time-scale for human natural genetic evolution seems much too grand for such developments to have any significant effect before other developments will have made the issue moot."
2.7. Technology Strategy
Bostrom has proposed specific strategies for technology development aimed at reducing existential risk. He advocates for the principle of differential technological development, which suggests that humanity should actively retard the development of technologies that increase existential risk while accelerating the development of beneficial technologies that protect against risks posed by nature or other technologies.
In 2011, he founded the Oxford Martin Program on the Impacts of Future Technology, further solidifying his commitment to strategic technological foresight. His theory of the unilateralist's curse has been cited as a reason for the scientific community to exercise extreme caution and avoid controversial and potentially dangerous research, such as the reanimation of pathogens, due to the risk of a single actor making a catastrophic decision that affects everyone.
3. Books and Writings
Nick Bostrom is a prolific author of books and academic articles that have significantly shaped discussions in philosophy, artificial intelligence, and futurology.
His major published books include:
- Anthropic Bias: Observation Selection Effects in Science and Philosophy (2002)
- Global Catastrophic Risks (2008), which he co-edited with Milan M. Ćirković
- Human Enhancement (2009), co-edited with Julian Savulescu
- Superintelligence: Paths, Dangers, Strategies (2014)
- Deep Utopia: Life and Meaning in a Solved World (2024)
Superintelligence: Paths, Dangers, Strategies became a New York Times Best Seller. The book argues for the possibility of superintelligence and explores its various types, cognitive abilities, and the associated risks. It also presents technical and strategic considerations for ensuring its safety.
In his 2024 book, Deep Utopia: Life and Meaning in a Solved World, Bostrom explores the concept of an ideal life in a hypothetical post-superintelligence world. He emphasizes that the crucial question is "not how interesting a future is to look at, but how good it is to live in." He outlines technologies that he considers theoretically possible at technological maturity, such as cognitive enhancement, reversal of aging, arbitrary sensory inputs (taste, sound), and precise control over motivation, mood, well-being, and personality. Bostrom suggests that in such a future, machines would not only outperform humans in labor but also undermine the purpose of many leisure activities, providing extreme welfare while simultaneously challenging humanity's quest for meaning.
Bostrom has also authored numerous influential journal articles, including:
- "How Long Before Superintelligence?" (1998)
- "Observer-relative chances in anthropic reasoning?" (2000)
- "The Meta-Newcomb Problem" (2001)
- "Existential Risks: Analyzing Human Extinction Scenarios and Related Hazards" (2002)
- "Are You Living in a Computer Simulation?" (2003)
- "The Mysteries of Self-Locating Belief and Anthropic Reasoning" (2003)
- "Astronomical Waste: The Opportunity Cost of Delayed Technological Development" (2003)
- "In Defense of Posthuman Dignity" (2005)
- "How Unlikely is a Doomsday Catastrophe?" (with Max Tegmark, 2005)
- "What is a Singleton?" (2006)
- "The Reversal Test: Eliminating Status Quo Bias in Applied Ethics" (with Toby Ord, 2006)
- "Converging Cognitive Enhancements" (with Anders Sandberg, 2006)
- "Drugs can be used to treat more than disease" (2008)
- "The doomsday argument" (2008)
- "Where Are They? Why I hope the search for extraterrestrial life finds nothing" (2008)
- "Letter from Utopia" (2008)
- "Cognitive Enhancement: Methods, Ethics, Regulatory Challenges" (with Anders Sandberg, 2009)
- "Pascal's Mugging" (2009)
- "Anthropic Shadow: Observation Selection Effects and Human Extinction Risks" (with Milan M. Ćirković and Anders Sandberg, 2010)
- "Information Hazards: A Typology of Potential Harms from Knowledge" (2011)
- "THE ETHICS OF ARTIFICIAL INTELLIGENCE" (2011)
- "Infinite Ethics" (2011)
- "The Superintelligent Will: Motivation and Instrumental Rationality in Advanced Artificial Agents" (2012)
- "Thinking Inside the Box: Controlling and Using Oracle AI" (with Stuart Armstrong and Anders Sandberg, 2012)
- "Existential Risk Reduction as Global Priority" (2013)
- "Embryo Selection for Cognitive Enhancement: Curiosity or Game-changer?" (with Carl Shulman, 2014)
- "Why we need friendly AI" (with Luke Muehlhauser, 2014)
- "The Vulnerable World Hypothesis" (2019)
4. Public Engagement and Evaluation
Bostrom has actively engaged with policy makers and the public, while his work has also drawn significant evaluation and, at times, controversy.
4.1. Public Engagement
Bostrom has provided policy advice and served as a consultant for numerous governments and international organizations. He has given evidence to the House of Lords Select Committee on Digital Skills in the UK Parliament. He is also an advisory board member for the Machine Intelligence Research Institute and the Future of Life Institute, and an external advisor for the Cambridge Centre for the Study of Existential Risk. Beyond formal advisory roles, Bostrom frequently delivers public lectures and appears in various media outlets, discussing topics such as cloning, artificial intelligence, mind uploading, cryonics, nanotechnology, and simulated reality, all within the broader context of transhumanism.
4.2. Criticism and Controversy
Bostrom's ideas, particularly his predictions regarding AI risk, have faced criticism. For instance, Oren Etzioni argued that the prediction of superintelligence being on a foreseeable horizon is not supported by available data. Similarly, Yann LeCun has asserted that there is no existential risk from superintelligent AI, contending that such AI would lack a desire for self-preservation and that experts can ensure its safety. Raffi Khatchadourian noted that Bostrom's book on superintelligence was not intended as a work of profound originality, but rather as an application of the rigorous methods of analytic philosophy to a complex body of ideas that originated at the periphery of academic thought. Despite some criticisms, Prospect magazine included Bostrom in its 2014 list of "World's Top Thinkers."
4.2.1. 1996 email incident
In January 2023, Nick Bostrom issued an apology for a 1996 email he sent to a listserv while he was a postgraduate student. In the email, he had stated his belief that "Blacks are more stupid than whites" and used a racial slur in describing how he thought this statement might be perceived by others. News sources widely described the email as "racist."
Bostrom's apology, posted on his website, acknowledged that "the invocation of a racial slur was repulsive" and stated that he "completely repudiate[d] this disgusting email." However, critics, including Andrew Anthony of The Guardian, noted that the apology did little to appease them, partly because Bostrom did not explicitly retract his central contention regarding race and intelligence and appeared to offer a partial defense of eugenics.
Following the incident, Oxford University condemned the language used in the email and initiated an investigation. The investigation concluded on August 10, 2023. The university stated that they did not consider Bostrom to be racist or to hold racist views, and that they found his January 2023 apology to be sincere. They also noted that they believed his apology, acknowledgment of the distress caused, and appreciation for the process were genuine, and that he had embarked on a journey of "deep and meaningful reflection." Despite this conclusion, the incident left a stain on the reputation of the Future of Humanity Institute, which Bostrom had championed.
5. Personal Life
Nick Bostrom met his wife, Susan, in 2002. As of 2015, Susan resided in Montreal, while Bostrom was based in Oxford. They have one son.
6. Influence
Nick Bostrom's research and philosophical thought have exerted significant influence across several academic and public spheres. His work has been particularly impactful in the fields of artificial intelligence, futurology, and existential risk studies, where he is recognized for bringing a rigorous philosophical approach to speculative yet critical topics.
He is widely credited for his foundational contributions to the discourse on longtermism, a philosophical stance that emphasizes the importance of positively influencing the long-term future. His advocacy has been instrumental in shaping the agenda for AI safety discussions, pushing for greater attention to the potential catastrophic risks associated with advanced artificial intelligence. By systematically analyzing the challenges and opportunities posed by future technologies, Bostrom has played a key role in establishing new areas of philosophical inquiry and encouraging interdisciplinary collaboration on issues vital to the future of humanity.
