I am now a research scientist at Google DeepMind.

Associate Professor in Computer Science
Maître de Conférences en Informatique

Dpt. MMIP, UMR AgroParisTech / INRA MIA 518

Address: AgroParisTech, 16 rue Claude Bernard, 75005 Paris

Phone#: 0033 1 44 08 16 79
Email: firstname.lastname@agroparistech.fr (Laurent ORSEAU) 

→ Teaching / Enseignement

Research

(fr) Un billet sur le HuffingtonPost vulgarisant quelques articles publiés avec Mark Ring.

(en) My main interest is Artificial General Intelligence (wikipedia), which overall goal is the grand goal of AI: building an autonomous machine which could be considered at least as intelligent as a human being.

Recently, the theory of this field has finally started to emerge, mainly thanks to some people like Ray Solomonoff,Marcus Hutter, and Jürgen Schmidhuber, who proposed very novel ideas and started to formalize the field. (Obviously, many others have contributed to this field!)

Hutter and Legg even gave a formal definition of Universal Intelligence (which you're allowed to debate, and I give my own here). The grand goal is far from achieved, and there remains a lot of work, but it's finally moving in the right direction.

Research interests (click on the links for details):

• Artificial intelligence
  • Artificial general intelligence
    • → Universal reinforcement learning and Solomonoff induction
    • → Mortal universal agents & wireheading
    • → Modifiable memory & space-time embedded intelligence
    • → Knowledge seeking agents and optimal scientists
    • → Teleporting intelligent agents
    • Self-modifiable agents
      1. Gödel machines, self-modifying universal agents
    • Levin search, OOPS, …
    • Incremental, continual, meta, transfer learning
  • Reinforcement learning
  • → Analogy making
  • Online, real-time learning
  • Neural networks
  • Minimum description length
• → Cellular automata

Publications

→ Click here

Organization of Events

• Co-organizer of EWRL11 - 2013 as the Dagstuhl Seminar on Reinforcement Learningwith Marcus Hutter and Peter Auer

• Co-organizer of the Uncertainty Handling workshop at AGI 2013 (at IJCAI) with Matthew Iklé

• Program Committee co-chair of AGI 2014 with Javier Snaider

Distinctions

• Kurzweil 2012 prize for the best idea for Artificial General Intelligence for the paper “Space-Time Embedded Intelligence” co-written with Mark Ring and presented at AGI 2012. See here for details.

• Solomonoff 2011 prize for the best contribution to the theory of Artificial General Intelligencefor two papers entitled “Self-Modification and Mortality in Artificial Agents” and “Delusion, Survival, and Intelligent Agents” co-written with Mark Ring and presented at AGI 2011. See here for details.

Misc

• Initiator of the Mathematical Artificial General Intelligence Consortium (MAGIC) mailing-list
• Member of the EWRL Steering Committee 2013-.
• Editor for Journal of AGI