Jump to content

Marcus Hutter

From Wikipedia, the free encyclopedia

Marcus Hutter
Portrait of Marcus Hutter
NationalityGerman
Alma materTechnical University Munich and Ludwig Maximilian University of Munich
Known forUniversal artificial intelligence
Artificial General Intelligence
AwardsIJCAI 2023
Alignment 2018
AGI 2016
UAI 2016
IJCAI-JAIR 2014
Kurzweil AGI 2009
Lindley 2006
Best Paper Prizes
Scientific career
Fields
InstitutionsDeepMind, Google, IDSIA, ANU, BrainLAB
ThesisInstantons in QCD (1996)
Doctoral advisorHarald Fritzsch
Other academic advisorsWilfried Brauer
Doctoral studentsShane Legg, Jan Leike and Tor Lattimore
Websitewww.hutter1.net

Marcus Hutter (born April 14, 1967 in Munich) is a professor and artificial intelligence researcher. As a senior researcher at DeepMind, he studies the mathematical foundations of artificial general intelligence.[1][2]

Hutter studied physics and computer science at the Technical University of Munich. In 2000 he joined Jürgen Schmidhuber's group at the Dalle Molle Institute for Artificial Intelligence Research in Manno, Switzerland.[3][4] He developed a mathematical formalism of artificial general intelligence named AIXI. He has served as a professor at the College of Engineering, Computing and Cybernetics of the Australian National University in Canberra, Australia.[5]

Research

[edit]

Starting in 2000, Hutter developed and published a mathematical theory of artificial general intelligence, AIXI, based on idealised intelligent agents and reward-motivated reinforcement learning.[6][7][4] His book Universal Artificial Intelligence: Sequential Decisions Based on Algorithmic Probability was published in 2005 by Springer.[8] Also in 2005, Hutter published with his doctoral student Shane Legg an intelligence test for artificial intelligence devices.[9] In 2009, Hutter developed and published the theory of feature reinforcement learning.[10] In 2014, Lattimore and Hutter published an asymptotically optimal extension of the AIXI agent.[11]

In 2019, Hutter joined DeepMind, recruited by Shane Legg.[2] In 2022, he co-authored a paper arguing that "deploying a sufficiently advanced reinforcement learning agent would likely be incompatible with the continued survival of humanity".[12][13]

Hutter Prize

[edit]
Main article: Hutter Prize

In 2006, Hutter announced the Hutter Prize for Lossless Compression of Human Knowledge, with a total of €50,000 in prize money.[14][15] In 2020, Hutter raised the prize money for the Hutter Prize to €500,000.[4]

See also

[edit]

Published works

[edit]
  • Marcus Hutter (2002). "The Fastest and Shortest Algorithm for All Well-Defined Problems". International Journal of Foundations of Computer Science. 13 (3). World Scientific: 431–443. arXiv:cs/0206022. Bibcode:2002cs........6022H. doi:10.1142/S0129054102001199. S2CID 5496821.
  • Marcus Hutter (2005). Universal Artificial Intelligence: Sequential Decisions Based on Algorithmic Probability. Springer. ISBN 9783540221395.
  • Joel Veness, Kee Siong Ng, Marcus Hutter, William Uther and David Silver (2011). "A Monte-Carlo AIXI Approximation". Journal of Artificial Intelligence Research. 40. AAAI Press: 95–142. arXiv:0909.0801. doi:10.1613/jair.3125. S2CID 206618.{{cite journal}}: CS1 maint: multiple names: authors list (link)

References

[edit]
  1. ^ "HomePage of Marcus Hutter".
  2. ^ a b "TIME100 AI 2023: Shane Legg". Time. 7 September 2023. Retrieved 25 August 2024.
  3. ^ "Marcus Hutter". aiandsociety.org. 15 May 2017. Retrieved 25 August 2024.
  4. ^ a b c Sagar, Ram (7 April 2020). "Compress Data And Win Hutter Prize Worth Half A Million Euros". Analytics India Magazine. Retrieved 7 March 2024.
  5. ^ "Professor Marcus Hutter". Australian National University. Retrieved 25 August 2024.
  6. ^ Marcus Hutter (2000). A Theory of Universal Artificial Intelligence based on Algorithmic Complexity. arXiv:cs.AI/0004001. Bibcode:2000cs........4001H.
  7. ^ Hutter, Marcus (28 November 2013). "To create a super-intelligent machine, start with an equation". The Conversation. Retrieved 25 August 2024.
  8. ^ Marcus Hutter (2005). Universal Artificial Intelligence: Sequential Decisions Based on Algorithmic Probability. Berlin; Heidelberg; New York: Springer. ISBN 9783540221395.
  9. ^ Graham-Rowe, Duncan (12 August 2005). "IQ test for AI devices gets experts thinking". New Scientist. Retrieved 25 August 2024.
  10. ^ Marcus Hutter (2009). "Feature Reinforcement Learning: Part I. Unstructured MDPs" (PDF). Journal of Artificial General Intelligence. ISSN 1946-0163.
  11. ^ Tor Lattimore and Marcus Hutter (2014). "Bayesian Reinforcement Learning with Exploration" (PDF). Algorithmic Learning Theory. Proc. 25th International Conf. on Algorithmic Learning Theory ({ALT'14}). Lecture Notes in Computer Science. Vol. 8776. pp. 170–184. doi:10.1007/978-3-319-11662-4_13. hdl:1885/14709. ISBN 978-3-319-11661-7.
  12. ^ Hutter, Marcus; Cohen, Michael K. (24 October 2022). "The danger of advanced artificial intelligence controlling its own feedback". The Conversation. Retrieved 25 August 2024.
  13. ^ Cohen, Michael K.; Hutter, Marcus; Osborne, Michael A. (August 2022). "Advanced artificial agents intervene in the provision of reward". AI Magazine. 43 (3): 282–293. doi:10.1002/aaai.12064. ISSN 0738-4602.
  14. ^ Chiang, Ted (9 February 2023). "ChatGPT Is a Blurry JPEG of the Web". The New Yorker. ISSN 0028-792X. Retrieved 25 August 2024.
  15. ^ Marcus Hutter. "50'000€ Prize for Compressing Human Knowledge". hutter1.net. Retrieved 29 November 2016.
Retrieved from "https://en.wikipedia.org/w/index.php?title=Marcus_Hutter&oldid=1242131793"