„Iyad Rahwan“ – Versionsunterschied

aus Wikipedia, der freien Enzyklopädie
Zur Navigation springen Zur Suche springen
Versionsgeschichte interaktiv durchsuchen
[ungesichtete Version][ungesichtete Version]
← Zum vorherigen VersionsunterschiedZum nächsten Versionsunterschied →
Inhalt gelöscht Inhalt hinzugefügt
VisuellWikitext
Fix cite errors
K →‎Cooperating with Machines: arxivify URL / redundant url
Zeile 28: Zeile 28:


===Cooperating with Machines ===
===Cooperating with Machines ===
Together with Jacob Crandall and others, Rahwan studied human-machine cooperation by exploring how state-of-the-art [[reinforcement learning]] algorithms perform when playing [[Repeated game|repeated games]] against humans. The authors showed that providing a medium of communication can result in an algorithm learning to cooperate with its human partner faster and more effectively than a human in these strategic games.<ref>{{Cite journal|last=Crandall|first=Jacob W.|last2=Oudah|first2=Mayada|last3=Tennom|last4=Ishowo-Oloko|first4=Fatimah|last5=Abdallah|first5=Sherief|last6=Bonnefon|first6=Jean-François|last7=Cebrian|first7=Manuel|last8=Shariff|first8=Azim|last9=Goodrich|first9=Michael A.|date=2018-01-16|title=Cooperating with machines|url=https://doi.org/10.1038/s41467-017-02597-8|journal=Nature Communications|language=En|volume=9|issue=1|doi=10.1038/s41467-017-02597-8|issn=2041-1723|pmc=5770455|pmid=29339817}}</ref><ref>{{Cite journal|url=https://arxiv.org/abs/1703.06207|title=[arXiv:1703.06207] Cooperating with Machines|last=|first=|date=|website=|archive-url=|archive-date=|dead-url=|access-date=|arxiv=1703.06207}}</ref><ref>{{Cite web|url=https://www.technologyreview.com/s/603995/ai-can-beat-us-at-pokernow-lets-see-if-it-can-work-with-us/|title=AI Can Beat Us at Poker—Now Let’s See If It Can Work with Us - MIT Technology Review|last=|first=|date=|website=|archive-url=|archive-date=|dead-url=|access-date=}}</ref>
Together with Jacob Crandall and others, Rahwan studied human-machine cooperation by exploring how state-of-the-art [[reinforcement learning]] algorithms perform when playing [[repeated game]]s against humans. The authors showed that providing a medium of communication can result in an algorithm learning to cooperate with its human partner faster and more effectively than a human in these strategic games.<ref>{{Cite journal|last=Crandall|first=Jacob W.|last2=Oudah|first2=Mayada|last3=Tennom|last4=Ishowo-Oloko|first4=Fatimah|last5=Abdallah|first5=Sherief|last6=Bonnefon|first6=Jean-François|last7=Cebrian|first7=Manuel|last8=Shariff|first8=Azim|last9=Goodrich|first9=Michael A.|date=2018-01-16|title=Cooperating with machines|url=https://doi.org/10.1038/s41467-017-02597-8|journal=Nature Communications|language=En|volume=9|issue=1|doi=10.1038/s41467-017-02597-8|issn=2041-1723|pmc=5770455|pmid=29339817}}</ref><ref>{{cite arxiv|arxiv=1703.06207|title=[arXiv:1703.06207] Cooperating with Machines|last=|first=|date=|website=|archive-url=|archive-date=|dead-url=|access-date=|arxiv=1703.06207}}</ref><ref>{{Cite web|url=https://www.technologyreview.com/s/603995/ai-can-beat-us-at-pokernow-lets-see-if-it-can-work-with-us/|title=AI Can Beat Us at Poker—Now Let’s See If It Can Work with Us - MIT Technology Review|last=|first=|date=|website=|archive-url=|archive-date=|dead-url=|access-date=}}</ref>


== AI and the Future of Work ==
== AI and the Future of Work ==

Version vom 27. Mai 2018, 17:06 Uhr

Vorlage:Infobox scientistIyad Rahwan (Vorlage:Lang-ar), is a Syrian-Australian scientist. He is an associate professor of Media Arts & Sciences at the MIT Media Lab, and is the director and principal investigator of its Scalable Cooperation group.[1] Rahwan's work lies at the intersection of the computer and social sciences, where he has investigated topics in computational social science, collective intelligence, large-scale cooperation, and the social aspects of artificial intelligence.[2]

Biography

Rahwan was born in Aleppo, Syria. He earned an Information Systems PhD in 2005 from the University of Melbourne. As an assistant and then associate professor in Computing and Information Science at MIT-partnered Masdar Institute of Science and Technology, Rahwan investigated scalable social mobilization's possibilities, limits, and challenges in various contexts by analyzing data from the 2009 DARPA Network Challenge,[3][4] the DARPA Shredder Challenge 2011,[5][6] and the 2012 US State Department Tag Challenge.[7][8][9] In 2015, Rahwan started the Scalable Cooperation Group at the MIT Media Lab, where he is the AT&T Career Development Professor and an Associate Professor of Media Arts & Sciences,[10] as well as an affiliate faculty at the MIT Institute of Data, Systems and Society.[11]

Society-in-the-Loop

Rahwan coined the term Society-in-the-loop as a conceptual extension of Human-in-the-Loop systems.[12][13] Whereas HITL systems embed an individual's judgement into a narrowly defined control system, SITL is more about embedding the judgement of society as a whole in to system. He cites an AI that controls billions of self driving cars (and decides who is worth saving in certain cases), or a news filtering algorithm with the potential to influence the ideology of millions of citizens (that decides what content the users shall see). Rahwan highlights the importance of articulating ethics and social contracts in ways that machines can understand, towards building new governance algorithms.[14]

Morality and Machines

Ethics of Autonomous Vehicles

Rahwan is one of the first to consider the problem of self autonomous vehicles as an ethical dilemma. His 2016 paper, The Social Dilemma of Autonomous Vehicles, showed that people approved of utilitarian autonomous vehicles, and wanted others to purchase these vehicles, but they themselves would prefer to ride in an autonomous vehicle that protected its passenger at all costs, and would not use self-driving vehicles if utilitarianism was imposed on them by law. Thus the paper concludes the regulation of utilitarian algorithms could paradoxically increase casualties by driving by inadvertently postponing the adoption of a safer technology.[15] The paper spurred lots of coverage about the role of ethics in the creation of artificially intelligent driving systems.[16][17][18][19][20][21][22]

Moral Machine

Moral Machine[23] is an online platform that generates ethical dilemma scenarios faced by hypothetical autonomous machines, allowing visitors to assess the scenarios and vote on the most morally acceptable between two unavoidable harm outcomes. As of April 2017, the system has collected 28 million decisions from over 3 million visitors.[24] The presented scenarios are often variations of the trolley problem, and the information collected would be used for further research regarding the decisions that machine intelligence might have to make in the future.[25][26][27]

Cooperating with Machines

Together with Jacob Crandall and others, Rahwan studied human-machine cooperation by exploring how state-of-the-art reinforcement learning algorithms perform when playing repeated games against humans. The authors showed that providing a medium of communication can result in an algorithm learning to cooperate with its human partner faster and more effectively than a human in these strategic games.[28][29][30]

AI and the Future of Work

Together with his student Morgan Frank and collaborators, Rahwan explored the relationship between city size and the potential impact of Artificial Intelligence and automation on employment. They used a variety of estimates of the risk of automation of different jobs.[31][32] Their main finding is that smaller cities may experience greater impact due to automation.[33] Related work explores the polarization of the US labor market, due to the underlying polarized structure of workplace skills.

Other projects

The Tag Challenge

Rahwan led the winning team in the 2012 US State Department Tag Challenge, using crowdsourcing and a referral-incentivizing reward mechanism (similar to the one used in the 2009 DARPA Network Challenge) to locate individuals in European and American cities within 12 hours each, given only their photographic portraits.[34][35][36]

The Nightmare Machine

The Nightmare Machine,[37] developed under Rahwan's guidance, creates computer generated imagery powered by deep learning algorithms to learn from human feedback and generate a visual approximation of what humans might find "scary".[38][39]

References

Vorlage:Reflist

  1. ↑ Group Overview ‹ Scalable Cooperation – MIT Media Lab.
  2. ↑ Iyad Rahwan - TEDxCambridge.
  3. ↑ How Social Media Mobilizes Society - LiveScience.
  4. ↑ A. Rutherford, M. Cebrian, S. Dsouza, E. Moro, A. Pentland, and I. Rahwan (2013). Limits of Social Mobilization. Proceedings of the National Academy of Sciences, vol. 110 no. 16 pp. 6281-6286.
  5. ↑ How Crowdsourcing Turned On Me - Nautilus.
  6. ↑ , Aamena Alshamsi, Manuel Cebrian, Iyad Rahwan: N. Stefanovitch, A. Alshamsi, M. Cebrian, I. Rahwan (2014). Error and attack tolerance of collective problem solving: The DARPA Shredder Challenge. EPJ Data Science. vol 3, no 13, pages 1-27. In: EPJ Data Science. 3. Jahrgang, 2014, doi:10.1140/epjds/s13688-014-0013-1 (springer.com).
  7. ↑ Crowdsourcing in manhunts can work : Nature News & Comment.
  8. ↑ , Manuel Cebrian, Iyad Rahwan, Sohan Dsouza, James McInerney, Victor Naroditskiy, Matteo Venanzi, Nicholas R. Jennings, J. R. Delara, Eero Wahlstedt, Steven U. Miller: A. Rutherford et al (2013). Targeted social mobilization in a global manhunt. PLOS ONE 8 (9): e74628. In: PLoS ONE. 8. Jahrgang, Nr. 9, 2013, S. e74628, doi:10.1371/journal.pone.0074628, PMID 24098660, bibcode:2013PLoSO...874628R (plos.org).
  9. ↑ Iyad Rahwan, Sohan Dsouza, Alex Rutherford, Victor Naroditskiy, James McInerney, Matteo Venanzi, Nicholas R. Jennings, Manuel Cebrian: Global Manhunt Pushes the Limits of Social Mobilization. In: Computer. 46. Jahrgang, Nr. 4, April 2013, ISSN 0018-9162, S. 68–75, doi:10.1109/mc.2012.295 (amerikanisches Englisch, doi.org).
  10. ↑ Person Overview ‹ Iyad Rahwan – MIT Media Lab.
  11. ↑ Iyad Rahwan – IDSS.
  12. ↑ Society in the Loop Artificial Intelligence ».
  13. ↑ Iyad Rahwan: Society-in-the-loop: programming the algorithmic social contract. In: Ethics and Information Technology. 20. Jahrgang, Nr. 1, 1. März 2018, ISSN 1388-1957, S. 5–14, doi:10.1007/s10676-017-9430-8 (englisch, springer.com).
  14. ↑ Society-in-the-loop.
  15. ↑ J. F. Bonnefon, A. Shariff, I. Rahwan (2016). The Social Dilemma of Autonomous Vehicles. Science. 352(6293):1573-1576.
  16. ↑ World Forum discuses how self-driving cars will make life or death decisions.
  17. ↑ Should Your Driverless Car Hit a Pedestrian to Save Your Life - The New York Times.
  18. ↑ Whose Life Should Your Car Save? - The New York Times.
  19. ↑ TedxCambridge: The social dilemma of driverless cars.
  20. ↑ Save the driver or save the crowd? Scientists wonder how driverless cars will ‘choose’ - The Washington Post.
  21. ↑ Driverless Cars Pose Difficult Ethical Question - Time.com.
  22. ↑ Driverless car safety revolution could be scuppered by moral dilemma - The Independent.
  23. ↑ Moral Machine.
  24. ↑ AI & Society at the Berkman Center.
  25. ↑ Ethical dilemma on four wheels: How to decide when your self-driving car should kill you - LA Times.
  26. ↑ For driverless cars, a moral dilemma: Who lives or dies? - Associated Press.
  27. ↑ Ethical dilemma on four wheels: How to decide when your self-driving car should kill you.
  28. ↑ Jacob W. Crandall, Mayada Oudah, Tennom, Fatimah Ishowo-Oloko, Sherief Abdallah, Jean-François Bonnefon, Manuel Cebrian, Azim Shariff, Michael A. Goodrich: Cooperating with machines. In: Nature Communications. 9. Jahrgang, Nr. 1, 16. Januar 2018, ISSN 2041-1723, doi:10.1038/s41467-017-02597-8, PMID 29339817, PMC 5770455 (freier Volltext) – (englisch, doi.org).
  29. ↑ Vorlage:Cite arxiv
  30. ↑ AI Can Beat Us at Poker—Now Let’s See If It Can Work with Us - MIT Technology Review.
  31. ↑ OECD Social, Employment and Migration Working Papers. ISSN 1815-199X, doi:10.1787/1815199x (englisch, doi.org).
  32. ↑ Carl Benedikt Frey, Michael A. Osborne: The future of employment: How susceptible are jobs to computerisation? In: Technological Forecasting and Social Change. 114. Jahrgang, Januar 2017, ISSN 0040-1625, S. 254–280, doi:10.1016/j.techfore.2016.08.019 (doi.org).
  33. ↑ Morgan R. Frank, Lijun Sun, Manuel Cebrian, Hyejin Youn, Iyad Rahwan: Small cities face greater impact from automation. In: Journal of The Royal Society Interface. 15. Jahrgang, Nr. 139, 1. Februar 2018, ISSN 1742-5689, S. 20170946, doi:10.1098/rsif.2017.0946, PMID 29436514, PMC 5832739 (freier Volltext) – (englisch, royalsocietypublishing.org).
  34. ↑ Crowdsourcing in Manhunts Can Work - Scientific American.
  35. ↑ Nowhere to hide: The next manhunt will be crowdsourced - New Scientist.
  36. ↑ Six degrees of mobilisation - The Economist.
  37. ↑ THe Nightmare Machine.
  38. ↑ Researchers Build 'Nightmare Machine' : The Two-Way : NPR.
  39. ↑ Clinton, Trump, the White House too, terrifyingly transformed by MIT’s ‘Nightmare Machine’ - The Washington Post.
Abgerufen von „https://de.wikipedia.org/w/index.php?title=Iyad_Rahwan&oldid=198881692“