A Study on Human Extinction Caused by Technology and the Path to Self-Salvation
DOI:
https://doi.org/10.63808/ihf.v2i3.505Keywords:
Continued development of science and technology, Extinction of humanity by science and technology, Human extinction, Extinction within one hundred years, Extinction within this century, Frontier technologies, Logic of argumentation, Time assessment, The Great Unification of humanity, Awakening movement, Social reconstruction, Non-competitive society, Peaceful society, Friendly society, Equitably prosperous societyAbstract
Science and technology are a double-edged sword. They can benefit humanity, yet they can also destroy it. The stronger their capacity to benefit humanity, the greater their power to destroy it. Since the Industrial Revolution in the mid-eighteenth century, science and technology have developed exponentially, yet humanity’s rational attitude toward science and technology has remained in a primitive state. I term this phenomenon evolutionary imbalance. Today, high-risk technologies have pushed humanity to the dangerous brink of self-extinction.
Since entering university in 1979, the author have been engaged in systematic research on technological risks for nearly half a century. Several major conclusions and a series of minor ones have emerged from this research. This article elaborates on only two of the important conclusions: first, “the continued development of science and technology will certainly soon lead to the extinction of humanity—at most within two or three hundred years, and at the earliest within this century”; second, “to save humanity from extinction, it is necessary to realize the Great Unification of humanity and strictly control science and technology.”
In 2007, my 800,000-character systematic research monograph Saving Humanity was published. In that book, my research conclusion was that the emergence of extinction-level means would not take more than 200 years, and that the eruption of extinction forces would not exceed 300 years thereafter. At the same time, I repeatedly emphasized that this was an extremely conservative estimate and that the reality is far more severe. Therefore, my conclusion was that “if science and technology continue to develop as they do today, they will certainly lead to the extinction of humanity—at most within four or five hundred years, and at the earliest within one hundred years,” and I made it clear that “this is a definitive conclusion, not merely a possibility.” (Hu, 2007a)
As early as January 2007, when Saving Humanity was still in proof, I sent it, together with The Open Letter to 26 Leaders of Mankind, to 25 leaders of major powers as well as the Secretary-General of the United Nations, in order to present this research conclusion and convey the gravity of the issue. Today, less than twenty years later, the development of science and technology has proven even faster than I imagined at the time. In fact, not long afterward, in the second edition of Saving Humanity, I revised the time-frame for the extinction of humanity by science and technology to “at most within two or three hundred years, and at the earliest within this century.” (Hu, 2019) Although “this century” also falls within “a hundred years,” this phrasing conveys a stronger sense of urgency. So, if this research conclusion is valid, what issue today could be more urgent and more important than solving this one? (Hu, 2025)
Over the years, I have devoted myself to using various methods to make people take this issue seriously, but with little effect. This paper presents the method of argumentation behind this conclusion, as well as the solution and pathway for addressing this urgent problem. The former is presented in two parts. The first part, “The Argument That the Continued Development of Science and Technology Will Certainly Lead to the Extinction of Humanity,” adopts the method of “Extinction Path Analysis + Defense Limit Testing.” The foundation of this analysis rests on five axioms, which can also be regarded as common sense—since axioms are inherently infallible, and common sense is naturally easy to understand. The reasoning process proceeds step by step through five key dimensions, demonstrating the inevitable connection between the continued development of science and technology and human extinction. The second part, “A Time Assessment of the Extinction of Humanity by Science and Technology,” is a comprehensive analysis based on the general laws of technological development along the timeline, combined with the current state of development of today’s frontier technologies. The latter is addressed in the third part, “The Solution and Realization Pathway for the Overall Survival Crisis of Humanity.”
All of these ideas were already present in Saving Humanity nearly twenty years ago. This paper revisits them today, incorporates new developments in science and technology from recent years, and draws on the research achievements of relevant scholars. Its purpose is to provide theoretical support and a practical warning for humanity to face technological risks squarely and adopt urgent countermeasures. (Hu, 2018)
References
[1] AI Impacts. (2023). Thousands of AI authors on the future of AI. Journal of Artificial Intelligence Research, 84, Article 9.
[2] Armstrong, S. (2013). Risks and mitigation strategies for Oracle AI. Studies in Applied Philosophy, Epistemology and Rational Ethics, 5, 335–347. https://doi.org/10.1007/978-3-642-31674-6_25
[3] Armstrong, S., Bostrom, N., & Shulman, C. (2013). Racing to the precipice: A model of artificial intelligence development. Technical Report.
[4] BBC Radio 4. (2024). Today [Video]. YouTube. https://www.youtube.com/watch?v=gnr0otD4HD0
[5] Bostrom, N. (2002). Existential risks: Analyzing human extinction scenarios and related hazards. Journal of Evolution and Technology, 9(1).
[6] Bostrom, N. (2003). Astronomical waste: The opportunity cost of delayed technological development. Utilitas, 15(3), 308–314.
[7] Bostrom, N. (2011). Information hazards: A typology of potential harms from knowledge. Review of Contemporary Philosophy, 10, 44–79.
[8] Bostrom, N. (2012). The superintelligent will: Motivation and instrumental rationality in advanced artificial agents. Minds and Machines, 22(2).
[9] Bostrom, N. (2013). Existential risk prevention as global priority. Global Policy, 4, 15–31.
[10] Bostrom, N. (2019). The vulnerable world hypothesis. Utilitas, 31.
[11] Bostrom, N. (2025) Deep Utopia: Life and Meaning in a Solved World. Beijing: CITIC Press.
[12] Cotton-Barratt, O., & Ord, T. (2015). Existential risk and existential hope: Definitions. Future of Humanity Institute Technical Report.
[13] Dampier, W. C. (2020). A brief history of science. China Overseas Chinese Press.
[14] Hu, J. Q. (2007a). Saving Humanity (1st ed., Vol. 1: The extinction of humanity). Tongxin Publishing House.
[15] Hu, J. Q. (2007b). Saving Humanity (1st ed., Vol. 2: The path for humanity). Tongxin Publishing House.
[16] Hu, J. Q. (2007c). The greatest problem. Northeastern University Press.
[17] Hu, J. Q. (2008). On human extinction. Ha Fai Yi Publication Limited.
[18] Hu, J. Q. (2018). Saving Humanity. Canada: FriesenPress.
[19] Hu, J. Q. (2019). Saving Humanity (3rd ed.). South Ocean Publishing House.
[20] Hu, J. Q. (2025). The Twelfth Open Letter to Human Leaders.
[21] Joy, B. (2000). Research priorities for robust and beneficial AI. Wired, 2000(4).
[22] Metaculus. (2018). If a global catastrophe occurs, will it reduce the human population by 95% or more? https://www.metaculus.com/questions/9060/#comments
[23] Milmo, D. (2025). AI firms warned to calculate threat of super intelligence or risk it escaping human control. The Guardian.
[24] Ord, T. (2021). The Precipice: Existential risk and the future of humanity. CITIC Press.
[25] Ord, T., Hillerbrand, R., & Sandberg, A. (2010). Probing the improbable: Methodological challenges for risks with low probabilities and high stakes. Journal of Risk Research, 13(4), 191–205.
[26] Orseau, L., & Armstrong, S. (2016). Safely interruptible agents. Proceedings of the Thirty-Second Conference on Uncertainty in Artificial Intelligence.
[27] PowerfulJRE. (2025). Joe Rogan Experience #2281 - Elon Musk [Video]. YouTube. https://www.youtube.com/live/sSOxPJD-VNo
[28] Rees, M. (2003a). Our Final Century.
[29] Rees, M. (2003b). Our Final Hour: A scientist’s warning. Basic Books.
[30] Rees, M. (2020). On the Future: Prospects for humanity. Shanghai Jiao Tong University Press.
[31] Russell, S. (2014). Of myths and moonshine. Edge.org.
[32] Russell, S. (2015). Unifying logic and probability: Recent developments. Communications of the ACM, 58(7).
[33] Russell, S. (2020). Human Compatible: Artificial intelligence and the problem of control. CITIC Publishing Group.
[34] Russell, S., & Norvig, P. (2023). Artificial Intelligence: A modern approach. People’s Posts and Telecommunications Press.
[35] Russell, S., Dewey, D. & Tegmark, M. (2015). Research priorities for robust and beneficial artificial intelligence. AI Magazine, 36, 105-114. https://doi.org/10.1609/aimag.v36i4.2577
[36] Snyder-Beattie, A. E., Ord, T., & Bonsall, M. B. (2019). An upper bound for the background rate of human extinction. Scientific Reports, 9.
[37] Stavrianos, L. S. (2003). A global history: From prehistory to the 21st century. Shanghai Academy of Social Sciences Press.
[38] Tegmark, M. (2017). Our mathematical universe: My quest for the ultimate nature of reality. Zhejiang People’s Publishing House.
[39] Tegmark, M. (2018). Life 3.0: Being human in the age of artificial intelligence. Zhejiang Education Publishing House.
[40] Tegmark, M., & Bostrom, N. (2005). How unlikely is a doomsday catastrophe? Nature.
[41] Toynbee, A. (2001). A study of history.
[42] Wu, G. S. The Course of Science (2022). Changsha: Hunan Science and Technology Press.
[43] Ху, Цзяци. Спасение Человечества (2020). Russia: Eksmo.
[44] Yudkowsky, E. (2001). Creating Friendly AI 1.0: The analysis and design of benevolent goal architectures. Machine Intelligence Research Institute.
[45] Yudkowsky, E. (2008). Cognitive biases potentially affecting judgement of global risks. In Global Catastrophic Risks (pp. 91–119).
[46] Yudkowsky, E. (2016). The AI alignment problem: Why it’s hard, and where to start. Machine Intelligence Research Institute.
[47] Yudkowsky, E. (2022). AGI Ruin: A list of lethality. https://intelligence.org/2022/06/10/agi-ruin/
[48] Yudkowsky, E., & Soares, S. (2025). If Anyone Builds It, Everyone Dies: Why superhuman AI would kill us all. Little, Brown and Company.
Downloads
Published
Issue
Section
License
Copyright (c) 2026 Jiaqi Hu

This work is licensed under a Creative Commons Attribution 4.0 International License.