Max Erik Tegmark (born 5 May 1967) is a Swedish-American physicist, machine learning researcher, and author. He is a professor of physics at the Massachusetts Institute of Technology and co-founder and president of the Future of Life Institute, a nonprofit focused on reducing global catastrophic risks from advanced technologies. Originally a cosmologist, Tegmark's research shifted toward artificial intelligence (AI) in the 2010s, and he has become a prominent voice advocating for careful management of AI development.
Tegmark's 2017 book Life 3.0 explores possible futures for humanity as AI advances, and he has called for a halt on the development of artificial superintelligence. His work spans precision cosmology, the mathematical universe hypothesis, and novel neural network architectures. He has received fellowships and awards for his contributions to physics and has been a central figure in public debates about AI safety.
Early life and education
Max Erik Tegmark was born Max Erik Shapiro in Stockholm, Sweden, on 5 May 1967, to Karin Tegmark and mathematician Harold S. Shapiro. He grew up in Bromma, Stockholm, and during his studies at Blackeberg's high school, he volunteered for the Swedish Peace and Arbitration Society, campaigning for nuclear disarmament. While in high school, he and a friend, Magnus Bodin, created and sold a word processor called Teddy, written in machine code for the Swedish eight-bit computer ABC 80, marketed by Liber Läromedel. He also coded a 3D Tetris-like game called Frac.
Tegmark left Sweden after receiving a B.A. in economics from the Stockholm School of Economics in 1989 and an M.S.E. in engineering physics from the KTH Royal Institute of Technology in 1990. He then studied physics at the University of California, Berkeley, earning an M.A. in 1992 and a Ph.D. in 1994 under the supervision of Joseph Silk. While at Berkeley, he adopted his mother's surname Tegmark, as there were many astronomers named Shapiro, including one of his professors.
Academic career
Tegmark began his academic career as an assistant professor at the University of Pennsylvania, where he received tenure in 2003. In 2004, he joined the Department of Physics at the Massachusetts Institute of Technology, where he has since been a professor. His early research focused on cosmology, but he later expanded into machine learning and AI safety. He has supervised numerous graduate students and postdoctoral researchers, contributing to a wide range of topics from cosmic microwave background analysis to interpretable neural networks.
Cosmology research
Tegmark has made significant contributions to precision cosmology, combining theoretical work with observational data to constrain cosmological models. He developed data analysis tools based on information theory and applied them to cosmic microwave background experiments such as COBE, QMAP, and WMAP, as well as galaxy redshift surveys including the Las Campanas Redshift Survey, the 2dF Survey, and the Sloan Digital Sky Survey.
In 1998, alongside Daniel Eisenstein and Wayne Hu, he introduced the idea of using baryon acoustic oscillations as a standard ruler in a paper published in The Astrophysical Journal. His 2000 paper in Physical Review E on quantum decoherence of neurons concluded that decoherence is too rapid for Roger Penrose's orchestrated objective reduction model of consciousness to be viable. In 2003, working with Angelica de Oliveira-Costa and Andrew Hamilton, Tegmark reported the discovery of the anomalous multipole alignment in the WMAP data, sometimes called the "axis of evil."
Tegmark also proposed the mathematical universe hypothesis in 2007, which postulates that the physical universe is a mathematical structure. This idea has been controversial; mathematician Edward Frenkel characterized it as "science fiction and mysticism" rather than science.
Machine learning and AI safety
In the 2010s, after about 25 years in cosmology and quantum information, Tegmark shifted his research focus to machine learning and AI safety. At MIT, he has worked on using AI to advance physics and on improving AI with insights from physics. In 2024, he co-authored a paper introducing Kolmogorov-Arnold Networks (KANs), which differ fundamentally from typical neural networks and are designed to be more interpretable. KANs are based on the Kolmogorov-Arnold representation theorem, which was previously thought to be irrelevant to machine learning. This work has implications for Deep learning and Neural network research, offering a potential alternative to standard architectures.
Tegmark has also engaged with broader AI topics, including Large language model safety and the societal impacts of Artificial intelligence. He has participated in public debates and policy discussions, often emphasizing the need for proactive risk mitigation.
Future of Life Institute
Tegmark co-founded the Future of Life Institute (FLI) with Anthony Aguirre, a professor at the University of California, Santa Cruz. The institute's mission is to "steer transformative technologies away from extreme, large-scale risks and towards benefiting life." It focuses on research to mitigate large-scale risks, particularly those from advanced AI. The board has included Elon Musk, Skype and Kazaa founder Jaan Tallinn, actors Alan Alda and Morgan Freeman, and individual graduate students, including Tegmark's wife, Meia Chita-Tegmark.
FLI has received significant funding, including a $665 million cryptocurrency donation from Vitalik Buterin in 2021, a $10 million donation from Elon Musk in 2015, and additional funding from Jaan Tallinn. Under Tegmark's leadership, FLI has supported research on AI safety and published open letters calling for caution in AI development.
In 2023, Tegmark was involved in a controversy when he was alleged to have signed a letter of intent for a $100,000 grant to far-right media outlet Nya Dagbladet, which was ultimately rejected. Tegmark stated that the institute rejected the grant after due diligence and that it finds Nazi, neo-Nazi, or pro-Nazi groups despicable. The institute issued a formal response denying any intentional support for such groups.
Journalism and public engagement
In 2020, Tegmark led a research project at MIT focused on using machine learning to classify news reports, called "Improving the News." To scale the work, he and his wife Meia Chita-Tegmark founded the Improve the News Foundation (ITN), an apolitical 501(c)(3) nonprofit, in October 2020. The product was rebranded as "Verity News" in 2023. This initiative aims to help readers understand complex issues in a nuanced way, using AI to reduce bias and misinformation.
Tegmark has also written for popular audiences, including his 2017 book Life 3.0, which has been translated into multiple languages. He frequently gives talks and interviews about AI risks and the future of humanity.
Awards and recognition
Tegmark was elected a Fellow of the American Physical Society in 2012 for his contributions to cosmology. He has received other honors for his research and public advocacy. His work has been widely cited, and he is a frequent speaker at conferences on AI and physics.
Personal life
Tegmark is married to Meia Chita-Tegmark, a cognitive scientist and AI researcher. They have collaborated on projects related to AI and media. Tegmark holds dual Swedish and American citizenship and has lived in the United States for most of his career.
See also
References
This article is based on publicly available information and source facts provided from Wikipedia (CC BY-SA).