“Humanity will become extinct in three years.”

The work we present here suggests that if the current decline in birth rates continues, humans could be extinct by 2394. … We also examined a scenario in which births ended in 2024, which revealed that Homo sapiens would become extinct in 2134.… Neither human nor animal subjects were used. Studies that predict species extinction have focused on a range of flora and fauna but in regard to Homo sapiens there are, with one notable exception, no predictive studies, only considerations of possible ways this may occur. The exception believes extinction of Homo sapiens will happen in 10,000 years. We agree that extinction will happen, but we disagree on the timing: The work we present here suggests that if the current decline in birth rates continues, humans could be extinct by 2394. If we consider the absence of working-age people and the accompanying collapse of services, the survivorship rates would most likely be lower. Given this, it is plausible that extinction could occur around 2359. We also examined a scenario in which births ended in 2024, which revealed that Homo sapiens would become extinct in 2134. Given societal collapse, extinction under the zero births scenario could occur around 2089. Method, Genetic Diversity, Natural Economy.

The world population is currently 8.1 billion, Given the decline in fertility between 2019 and 2024 and employing a probabilistic forecasting method in conjunction with 66% confidence intervals, we find: by 2139 the world population will be between 1.55 billion and 1.81 billion; by 2239, it will be between 4.95 million and 5.84 million and by 2339 there will be no humans. … Moreover, we agree with Gee that extinction will happen, but disagree on the timing: The work we present here suggests that humans could be extinct in 314 years, well before 10,000 years have passed, when one takes uncertainty into consideration## Abstract Studies that predict species extinction have focused on a range of flora and fauna but in regard to Homo sapiens there are, with one notable exception, no predictive studies, only considerations of possible ways this may occur. The exception believes extinction of Homo sapiens will happen in 10,000 years. We agree that extinction will happen, but we disagree on the timing. The world population is currently 8.1 billion, Given the decline in fertility between 2019 and 2024 and employing a probabilistic forecasting method in conjunction with 66% confidence intervals, we find: by 2139 the world population will be between 1.55 billion and 1.81 billion; by 2239, it will be between 4.95 million and 5.84 million and by 2339 there will be no humans. ## Introduction … and Swanson, 1999, 2025) , which includes long range forecasting (Swanson, 2019; Swanson, 2024; Tayman and Swanson, 2025b) , as well as experience with both probabilistic forecasting (Swanson and Beck, 1994; Swanson and Tayman, 2014; Swanson and Tayman, 2024, 2025b; Swanson, Tayman, and Cline, 2025) and the role of assumptions in forecasts (Smith, Tayman, and Swanson, 2001 : 286, 2013 : 308, Swanson and Tayman, 1995; Swanson, Schlottmann and Schmidt, 2009; Swanson, Tayman, and Cline, 2025 ) Moreover, we agree with Gee that extinction will happen, but disagree on the timing: The work we present here suggests that humans could be extinct in 314 years, well before 10,000 years have passed, when one takes uncertainty into consideration (Taleb 2010) , a factor only recently formally introduced to population projections in terms of the CCM and H-P methods (Alkema et al. 2015; Swanson, Tayman, and Cline, 2025; United Nations, 2024c ). . The population declines by -6.0% during the first 50 years of the projection horizon. That decline increases to -63.4% during the next 50 years. By 2124, the world's population is 2.8 billion, a drop of 5.3 billion people from 2024. …

Our simulation indicates rapid decline in food production and unequal distribution of ~6 billion deaths due to starvation by 2100. … Ultimately, following a period of continued growth to 2050, our results indicate increasing risk of mass mortality due to food insecurity with a decline to ~32 % of current population by 2100.… Yet, such speculative ‘runaway global warming’ has received minimal attention compared to mainstream low- to mid range scenarios. This study builds on The Limits to Growth to provide new insights into the in ternational risk of mass mortality due to food insecurity based on a higher-resolution illustration of World3’s ‘runaway global warming’ scenario (~8–12 ◦C+). Our simulation indicates rapid decline in food production and unequal distribution of ~6 billion deaths due to starvation by 2100. We highlight the importance of including high-resolution simulations of high-range global warming in climate change impact modelling to make well-informed decisions about climate change mitigation, resilience and adaptation. ## 1. Introduction … Sustained ‘fossil-fueled development’ (the narrative underpinning SSP5-RCP8.5 and World3) would see growth in population, wealth and high-demand lifestyles continue through the first half of the century. However, the resulting climate-induced agricultural failures coupled with international trade competition would see global food insecurity cascade. Ultimately, following a period of continued growth to 2050, our results indicate increasing risk of mass mortality due to food insecurity with a decline to ~32 % of current population by 2100. We observe some discrepancies between our simulation output and that of World3, given our less complex and less aggregated (i.e., gloval-scale-national-granularity) model compared with the more complex and more aggregated (i.e., global-scale-global-granularity) nature of World3. Fig. …

We find that, over the course of the 21st century, the Doomsday forecast is well approximated by an extinction risk of 0.2% per year. … Here we estimate a 13% chance that humanity shall fail to see out the 21st century. … Nonetheless, in this particular model, our civilisation’s ‘life expectancy’ (the date we are equally likely to pass or fall short of) is approximately 700 years from the present.… In order to achieve this we require a model for the future global birth rate. Over the past thirty years, this number has remained approximately constant, at close to 140 million per year [24]. In the short term we can therefore be confident that this number will serve as a good estimate, barring a global catastrophe. We find that, over the course of the 21st century, the Doomsday forecast is well approximated by an extinction risk of 0.2% per year. How do these results compare with other assessments of our future prospects? For example, Rees [14] estimated a 50% probability of human extinction by the year 2100. A study by Sandberg and Bostrom [25] arrived at a risk factor of 19% over the same timeframe. The results of this work are only slightly more optimistic. Here we estimate a 13% chance that humanity shall fail to see out the 21st century. This equates to odds of 7:1. Pressing further into the future, the forecast becomes less reliable, since the birth rate could change significantly. Nonetheless, in this particular model, our civilisation’s ‘life expectancy’ (the date we are equally likely to pass or fall short of) is approximately 700 years from the present. This appears consistent with an independent estimate of ∼ 400 years, which derives from the longevity of historical civilisations [26]. ## The Total Number of Observers

The short answer is that we do not know whether climate change or anthropogenic ecological disruption could spiral into human extinction. … Indeed, no single hazard is likely to result directly in human extinction. The search for one single event to kill us all will lead us to science fiction.… Few attempts have been made, with the notable exception of the societal collapse and climate review conducted by Richards et al., which does attempt to cautiously trace out some pathways from impacts to conflict and mass mortality. 65 Notably, these deal only with climate change and not the broader, reinforcing web of ecological crises, which has received less attention. The short answer is that we do not know whether climate change or anthropogenic ecological disruption could spiral into human extinction. However, this is true for all the suspected causes of human extinction. Climate and ecological crises do appear to have one of the most concerning profiles, given their range of impacts, as well as their role in past mass-extinction events and periods of historical turmoil. There are enough reasons to take this question of human extinction from ecological breakdown seriously. For now, while uncertainty remains, it seems improbable that human actions could extinguish the biosphere. … Whether our combined global environmental crises could spiral into extinction depends on human responses and wider trends and vulnerabilities (such as inequality). Climate change and planetary boundaries challenge the traditional, simplistic approach of thinking of existential risk as a simple set of disconnected hazards. Indeed, no single hazard is likely to result directly in human extinction. The search for one single event to kill us all will lead us to science fiction. 70 We should instead think of the overall level of risk that arises from any particular socio-economic system (such as the current fossil-fuel-driven, globalised, capitalist economy). Answering the question of whether climate change is an existential risk is a futile inquiry until we develop reasonable definitions of existential risk, a topic we turn to next. Limits to growth as an existential saviour and threat Can we grow into catastrophe, collapse, or even human extinction? …

In The Precipice, Toby Ord judges the overall likelihood of an existential catastrophe in the coming century at about one in six. … The literature suggests that the likelihood of human extinction from exogenous threats, where we have more robust data to draw on, is extremely low. For anthropogenic threats, there is far greater uncertainty.… A second category relates to entirely unprecedented developments, for which we must often primarily draw on experts’ subjective judgements. A third challenge relates to scenarios involving mul tiple interacting risks or trends; the complexity of these scenarios makes them challenging to assign likelihood to. Plainly put, humans are quite different from other species. We have no similar extinctions to draw on and so must fumble in the dark with what we have. In The Precipice, Toby Ord judges the overall likelihood of an existential catastrophe in the coming century at about one in six. This figure is itself informed by subjective judgements across a range of individual risks where certainty is not possible. Ord caveats the purpose of these judgements as showing “the right order of magnitude, rather than a more precise probability“; even so, it would be reasonable to regard them as incorporating some element of guesswork given the presently unknowable nature of many of the informing variables. … ## Conclusion The literature suggests that the likelihood of human extinction from exogenous threats, where we have more robust data to draw on, is extremely low. For anthropogenic threats, there is far greater uncertainty. For many anthropogenic threats such as nuclear war, pandemics and environmental degradation, the likelihood of such events and the scale of their catastrophic consequences strongly justify prevention and mitigation efforts. However, it is unclear how plausible they are as causes of extinction of homo sapiens. …

Under high-alert nuclear postures, estimated annual hazards range from roughly 0.1% to 2%, implying median survival times on the order of decades to millennia conditional on institutional configuration. … Across scenarios, median hazard estimates from the Bayesian scenario distributions lie in the range 10−3–10−2 per year, with wide but finite 90% credible intervals spanning roughly an order of magnitude within each scenario. … Together, they indicate that under high-alert nuclear configurations, detectable civilizational lifetimes plausibly lie in the range of decades to centuries, rather than millennia or longer.… A Bayesian survival-analysis framework with regime-switching hazards captures temporal non-stationarity between Cold War and post-Cold War periods and propagates parameter uncertainty. Under high-alert nuclear postures, estimated annual hazards range from roughly 0.1% to 2%, implying median survival times on the order of decades to millennia conditional on institutional configuration. These hazards plausibly exceed natural extinction risks but may decline substantially with improved governance and control systems. The results bound detectable civilizational longevity and offer conditional implications for existential-risk prioritization and SETI expectations under deep uncertainty. ## 1 Introduction … ### 5.2 Bayesian Posterior Hazard Distributions Using (MCMC) sampling of the Bayesian scenario model described, we obtain scenario distributions for λCW and λpost by propagating uncertainty in: * Escalation probability (p*esc*), * Catalog completeness (α), * Regime reduction parameter (r). Across scenarios, median hazard estimates from the Bayesian scenario distributions lie in the range 10−3–10−2 per year, with wide but finite 90% credible intervals spanning roughly an order of magnitude within each scenario. Two features are noteworthy. First, posterior medians are generally close to the deterministic benchmarks but not identical. Because λ depends multiplicatively on uncertain parameters, the scenario distribution is right-skewed, and Jensen-type effects shift the median relative to plug-in values. … The deterministic plug-in benchmarks provide transparent central-case illustrations. The Bayesian posterior generalizes these by incorporating uncertainty and revealing skewed survival distributions. Together, they indicate that under high-alert nuclear configurations, detectable civilizational lifetimes plausibly lie in the range of decades to centuries, rather than millennia or longer. These results remain conditional on Earth-like institutional and technological structures. They do not imply that nuclear war is the Great Filter, nor that most civilizations follow Earth's trajectory. …

We need to distinguish between climate doomists (who see catastrophe as imminent and unavoidable) and climate risk realists (who see catastrophe as one potential future that should be avoided).## Abstract Many now argue that we should think about the previously unthinkable risks of climate change, including societal collapses and human extinction. Calami tous images of the future are not pathological or counterproductive: it is both necessary and valuable to imagine the worst-case scenarios of climate change. Critics of climate catastrophe often group together all visions of disastrous futures under labels like doomism or pessimism. This is unhelpful and greater nuance is required. We need to distinguish between climate doomists (who see catastrophe as imminent and unavoidable) and climate risk realists (who see catastrophe as one potential future that should be avoided). We also need to split apart the different ways of envisioning climate catastrophe to understand their distinct strengths and weaknesses. We outline and compare three alterna tive modes of viewing the worst-case scenarios of climate change: foresight, agitation, and fiction. The first centers on modeling catastrophic climate sce narios, the second on the use of images of climate catastrophe for political action, and the third on fictional visions of future climate disasters. …

An ex-Anthropic researcher claims AI could kill us all by 2030. … “The people building AI earnestly believe that it could kill us all by the end of the decade,” Coxon said.The ex-Anthropic researcher's warning that AI could kill us all is missing something important | Fortune # An ex-Anthropic researcher claims AI could kill us all by 2030. But he fails to answer the most essential question: What are we supposed to do about it? September 10, 2026, 4:24 PM ET … > Another day, another doomsday AI post that leaves the public feeling unsettled, anxious, and unsure what to do next. At least that’s how it feels as an AI reporter who reports on this space every day. This time, the alarm bells are coming from inside the industry, courtesy of a former researcher at OpenAI and Anthropic named Jacob Coxon. His X post on Sept. 8 has gone mega-viral, racking up over 150 million views as of this writing. “The people building AI earnestly believe that it could kill us all by the end of the decade,” Coxon said. “I spent the last three years doing pre-training research at both OpenAI and Anthropic. Neither company is acting responsibly. They are racing straight to self-improving superintelligence and gambling with our lives.” Anthropic’s current head of alignment re-posted Coxon, noting that he is “correct” that “we really do earnestly believe AI could kill all humans!” He estimates there’s a greater than 10% chance it could happen within the next decade. …

The median expert put 6 percent odds that humans will go extinct by 2100; they estimated 20 percent odds of a catastrophic event before the year 2100 that kills off at least 10 percent of the human population within a five year period. … The superforecasters, by contrast, are more optimistic, putting a 9 percent chance of catastrophe and a 1 percent chance of extinction.… To do this, they asked subject matter experts who study threats that at least conceivably could jeopardize humanity’s survival (like nuclear weapons, pandemics, climate change, and rogue artificial intelligence), but they also asked superforecasters who’ve proven accurate at predicting events in the past. The superforecaster group is not made up of experts on existential threats to humanity, but rather generalists from a variety of occupations with solid predictive track records. The median expert put 6 percent odds that humans will go extinct by 2100; they estimated 20 percent odds of a catastrophic event before the year 2100 that kills off at least 10 percent of the human population within a five year period. (To put into perspective just how catastrophic such a catastrophic event would be, World War II resulted in the deaths of less than 4 percent of the global population at the time.) The superforecasters, by contrast, are more optimistic, putting a 9 percent chance of catastrophe and a 1 percent chance of extinction. These are astonishingly large risks. The expert survey suggests that humanity has worse odds of surviving to 2100 than a man diagnosed with prostate cancer has of living another five years; the superforecasters estimate that humans are likelier to go extinct than an average person is to be audited by the IRS.

Nuclear war, when sampling from an artificial normal distribution, results in an average human survival time of 60 years into the future starting from the present, before a civilization-ending disaster. While climate change results in an average human survival time of 193 years, the simulation based on impact from asteroids results in an average of 1754 years.… providing a direction to approach this problem. Using a probabilistic model, we observed the effects of nuclear war, climate change, asteroid impacts, artificial intelligence and pandemics, which are the most harmful disasters in terms of their extent of destruction on the length of human survival. We consider the starting point of the predicted average number of survival years as the present calendar year. Nuclear war, when sampling from an artificial normal distribution, results in an average human survival time of 60 years into the future starting from the present, before a civilization-ending disaster. While climate change results in an average human survival time of 193 years, the simulation based on impact from asteroids results in an average of 1754 years. Since the risks from asteroid impacts could be considered to reside mostly in the far future, it can be concluded that nuclear war, climate change, and pandemics are presently the most prominent threats to humanity. Additionally, the danger from superiority of artificial intelligence over humans, although still somewhat abstract, is worthy of further study as its potential for impeding humankind’s progress towards becoming a more advanced civilization cannot be confidently dismissed. 1. …

In other words, given that there have been 200,000 years with no human extinction event, the probability of humanity going extinct in any given year is approximately 2x10 -6.… (2019) briefly considers the interplay between natural GCRs and artificial systems; Ord (2020) does both in greater detail. 2 Equation (1) corresponds to Equation 9 of Bailey (1997). Equation 20 of Quigley and Revie (2011) is similar, using 2.5 instead of 3 in the denominator. In Equation (1), n is the number of time periods that have elapsed without a failure event. Thus, for n=200,000, it follows that P≈2x10 -6. In other words, given that there have been 200,000 years with no human extinction event, the probability of humanity going extinct in any given year is approximately 2x10 -6. That is a rather small number, so it may be helpful to consider the probability per millennium instead of per year. Given that there have been 200 millennia with no extinction event, the probability of humanity going extinct in any given millennium is approximately 0.002, or 0.2%. That is indeed a small probability. A caveat is that our ability to do this analysis depends on humanity’s continued existence. This …

In my view, the chance that AI will eliminate humans in the next five years is all but indistinguishable from zero. Humans are too geographically spread out, too genetically diverse, and too resourceful to simply fall apart altogether. The idea that AI will kill us all in five years is preposterous.Certainly Coxon has not (thus far) added anything substantive exto the argument. In my view, the chance that AI will eliminate humans in the next five years is all but indistinguishable from zero. Humans are too geographically spread out, too genetically diverse, and too resourceful to simply fall apart altogether. The idea that AI will kill us all in five years is preposterous. And no, to answer a comment I got on X when I raised these issues, an AI-generated virus is not likely to kill literally all of us, either. When I asked the virologist Dr. Angela Rasmussen about this, she told me “i mean, never say never but the only virus I know that has near total mortality is rabies, and considering it’s not airborne and takes 2-4 weeks to kill you, I don’t think it would take out all of humanity. Ebola doesn’t have 90% mortality like the first outbreak did. …

• Existential risks are events that, either directly or indirectly, could cause the extinction of humanity or the irreversible collapse of society worldwide. 11… If unchecked, GCRs could lead to great suffering, loss of life, and sustained damage to national governments, international relations, economies, societal stability and global security (generalized from a definition of global catastrophic biological risks). 9 GCRs could result in the loss of at least 10 per cent of the global population or the alteration of the future trajectory of humanity, but from which recovery is possible. 10 • Existential risks are events that, either directly or indirectly, could cause the extinction of humanity or the irreversible collapse of society worldwide. 11 The 10 per cent threshold in the GCR definition is an arbitrary value, but one that has been widely adopted within the field of existential risk as a kind of shorthand representing the scale and scope envisioned. …

Professor Hinton, who said the pace of technological change is faster than expected, diagnosed that there is a 10-20% chance that humans will become extinct within the next 30 years due to AI.The Guardian reported on the 27th (local time) that Professor Hinton appeared on BBC radio and feared that the rapid development of AI technology could pose a threat to humanity's survival. Professor Hinton, who said the pace of technological change is faster than expected, diagnosed that there is a 10-20% chance that humans will become extinct within the next 30 years due to AI. In the past, he has argued that there is a 10% chance that technological advances will lead to catastrophic consequences for mankind.

The scientific consensus is that there is a relatively low risk of near-term human extinction due to natural causes. [2] [3] The likelihood of human extinction through humankind's own activities, however, is a current area of research and debate.Some of the many possible contributors to anthropogenic hazards are climate change, global nuclear annihilation, biological warfare, weapons of mass destruction, and ecological collapse. Other scenarios center on emerging technologies, such as advanced artificial intelligence, biotechnology, or self-replicating nanobots. The scientific consensus is that there is a relatively low risk of near-term human extinction due to natural causes. [2] [3] The likelihood of human extinction through humankind's own activities, however, is a current area of research and debate. ### Early history

Marks made five: that AI developers believe their technology could cause human extinction, possibly within a few years, and that seniority correlates with concern;### Five numbered claims Marks made five: that AI developers believe their technology could cause human extinction, possibly within a few years, and that seniority correlates with concern; that they continue because of commercial incentives plus a belief they are racing less responsible rivals; that AIs cannot be programmed like traditional software and “frequently severely misbehave”; that no method robustly aligns them; and that many staff want to slow down.

XPT superforecasters predicted that catastrophic and extinction risk from AI by 2030 is very low (0.01% catastrophic risk and 0.0001% extinction risk).## AI risk - XPT superforecasters predicted that catastrophic and extinction risk from AI by 2030 is very low (0.01% catastrophic risk and 0.0001% extinction risk). - XPT superforecasters predicted that catastrophic risk from nuclear weapons by 2100 is almost twice as likely as catastrophic risk from AI by 2100 (4% vs 2.13%). - XPT superforecasters predicted that extinction risk from AI by 2050 and 2100 is roughly an order of magnitude larger than extinction risk from nuclear weapons, which in turn is an order of magnitude larger than non-anthropogenic extinction risk (see here for details). …

What will be the fate of human civilization millions, billions, or trillions of years into the future? This paper, written by an international group of 14 scholars and published in the journal Foresight, studies several possible long-term trajectories of human civilization, including the radically good and catastrophically bad.#### Long-Term Trajectories of Human Civilization What will be the fate of human civilization millions, billions, or trillions of years into the future? This paper, written by an international group of 14 scholars and published in the journal Foresight, studies several possible long-term trajectories of human civilization, including the radically good and catastrophically bad. #### A Model for the Probability of Nuclear War

In 2022, a survey of AI researchers with a 17% response rate found that the majority believed there is a 10 percent or greater chance that human inability to control AI will cause an existential catastrophe.… Debates center on AGI's technical feasibility, the speed of self-improvement, [7] and the effectiveness of alignment strategies. [8] Concerns about superintelligence have been voiced by researchers including Geoffrey Hinton, [9] Yoshua Bengio, [10] Demis Hassabis, [11] and Alan Turing, [a] and AI company CEOs such as Dario Amodei (Anthropic), [14] Sam Altman (OpenAI), [15] and Elon Musk (xAI). [16] In 2022, a survey of AI researchers with a 17% response rate found that the majority believed there is a 10 percent or greater chance that human inability to control AI will cause an existential catastrophe. [17] [18] In 2023, hundreds of AI experts and other notable figures signed a statement declaring, "Mitigating the risk of extinction from AI should be a global priority alongside other societal-scale risks such as pandemics and nuclear war". [19] Following increased concern over AI risks, government leaders such as United Kingdom prime minister Rishi Sunak [20] and United Nations Secretary-General António Guterres [21] called for an increased focus on global AI regulation. …

In total, experts predicted a 6% chance of human extinction by 2100. Superforecasters put the odds at 1%.Credit: The Forecasting Research Institute Although predictions ranged widely and there was disagreement within the groups, the experts were generally more concerned about all kinds of existential risks. In total, experts predicted a 6% chance of human extinction by 2100. Superforecasters put the odds at 1%. Experts predicted a 20% chance of a catastrophe this century, while superforecasters estimated less than half of that (around 9%). Credit: The Forecasting Research Institute

The question is not whether humans will become extinct, but rather when they will. If the anti-natalist arguments are correct, it would be better, all things being equal, if this happened sooner rather than later for, the sooner it happens, the more suffering and misfortune will be avoided.Rearing children, whether one’s biological offspring or adopted, can bring satisfaction. If the number of unwanted children were to ever come to zero, anti-natalism would entail the deprivation of this benefit to those who accept the moral prohibition on creating children. That does not mean that we should reject anti-natalism. The reward of becoming a parent does not outweigh the serious harm procreation will cause to others. The question is not whether humans will become extinct, but rather when they will. If the anti-natalist arguments are correct, it would be better, all things being equal, if this happened sooner rather than later for, the sooner it happens, the more suffering and misfortune will be avoided.

For instance, if bees go extinct, humans will become extinct in the next 10 years.Philanthropy, by definition, is the love for human beings. As the planet faces an environmental crisis, the survival of our species is under threat. For us to survive, we need to protect other species because we survive symbiotically and not in isolation. Biodiversity is therefore crucial to preserving ourselves. It is not just about plants and trees; it involves a fragile balance of hundreds of co-dependent species. For instance, if bees go extinct, humans will become extinct in the next 10 years. As a practising pantheist, who believes that nature is god, it is my dharma to conserve the ecosystem unique to this region.