<div dir="ltr">(Apologies for multiple postings)<br><div class="gmail_quote"><div dir="ltr"><br clear="all"><div>What: <span>MAchine</span> <span>INtelligence</span> workshop at NIPS 2016 (MAIN@NIPS)<br>When and Where: Fri Dec 9th 2016, 9am-12.30pm and 2pm-5.30pm<br>Where: Centre Convencions Internacional, Barcelona (Spain) (co-located with NIPS)<br>Website: <a href="https://mainatnips.github.io/" target="_blank">https://mainatnips.github.io/</a><br><br></div><div>ABSTRACT SUBMISSION DEADLINE APPROACHING: Sunday, October 9th</div><div><br>*** Motivation ***<br><br>Recent years have seen the success of <span>machine</span>
learning systems, in particular deep learning architectures, on
specific challenges such as image classification and playing Go.
Nevertheless, <span>machines</span> still fail on hallmarks of human <span>intelligence</span>
such as the flexibility to quickly switch between a number of different
tasks, the ability to creatively combine previously acquired skills in
order to perform a more complex goal, the capacity to learn a new skill
from just a few examples, or the use of communication and interaction to
extend one's knowledge in order to accomplish new goals.<br><br>This workshop aims to stimulate theoretical and practical advances in the development of <span>machines</span> endowed with human-like general-purpose <span>intelligence</span>, focusing in particular on benchmarks to train and evaluate progress in <span>machine</span> <span>intelligence</span>. The workshop will feature invited talks by top researchers from <span>machine</span>
learning, AI, cognitive science and NLP, who will discuss with the
audience their ideas about what are the most pressing issues we face in
developing true AI and the best methods to measure genuine progress. We
are moreover calling for position statements from interested researchers
to complement the workshop program (see below).<br><br>The workshop
will also introduce the new Environment for Communication-Based AI (CommAI-env) to
the research community, encouraging discussion on how to make it the
ultimate benchmark for <span>communication-based machine</span> <span>intelligence</span>. The Environment aims at being an interactive playground where systems can only succeed if they possess the hallmarks of <span>intelligence</span>
we listed above. A prototype of the environment is available here: <a href="https://github.com/facebookresearch/CommAI-env" target="_blank">https://github.com/<wbr>facebookresearch/CommAI-env</a><br><br>*** Call for abstracts ***<br><br>We
invite submission of 3-page abstracts (or shorter) presenting a
position statement on what are the most important challenges to develop
general <span>machine</span> <span>intelligence</span> (with special interest in communication-based <span>intelligence</span>),
and what are the data and benchmarks we need to properly address such
challenges. While all points of view are welcome, we also encourage critiques of the Environment for Communication-Based AI. Abstracts must be submitted in PDF format to <a href="mailto:main.at.nips2016@gmail.com" target="_blank">main.at.nips2016@gmail.com</a>
by Sunday, October 9th. We will notify acceptance by Sunday, October
23d. Authors of accepted abstracts are expected to present their ideas
in an oral presentation at the workshop, and to take part in the panel
discussions.<br><br>*** Invited speakers ***<br><br>* Emannuel Dupoux (Laboratoire de Sciences Cognitives et Psycholinguistique, Paris): <a href="http://www.lscp.net/persons/dupoux/" target="_blank">http://www.lscp.net/persons/du<wbr>poux/</a><br><br>* Fernando Diaz (Microsoft Research, New York): <a href="http://msr.nyc/fdiaz/" target="_blank">http://msr.nyc/fdiaz/</a><br><br>* Raquel Fernandez (Institute for Logic, Language & Computation, Amsterdam): <a href="https://staff.fnwi.uva.nl/r.fernandezrovira/" target="_blank">https://staff.fnwi.uva.nl/r.fe<wbr>rnandezrovira/</a><br><br></div><div>* Brenden Lake (NYU Center for Data Science): <a href="http://cims.nyu.edu/~brenden/" target="_blank">http://cims.nyu.edu/~brenden/</a><br></div><div><br>* Jürgen Schmidhuber (IDSIA, Lugano): <a href="http://people.idsia.ch/~juergen/" target="_blank">http://people.idsia.ch/~<wbr>juergen/</a><br></div><div><br>* Arthur Szlam (Facebook Artificial <span>Intelligence</span> Reserch, New York): <a href="https://research.facebook.com/arthur-szlam" target="_blank">https://research.facebook.com/<wbr>arthur-szlam</a><br><br>* Julian Togelius (NYU Game Innovation Lab, New York): <a href="http://julian.togelius.com/" target="_blank">http://julian.togelius.com/</a><br><br>*** Organizers ***<br><br>Tomas Mikolov, Allan Jabri, Armand Joulin, Klemen Simonic (Facebook)<br>Marco Baroni, Angeliki Lazaridou, Germán Kruszewski (University of Trento/Facebook)<br><br>*** Program ***<br><br>The workshop will feature oral presentations and two panel discussions. The program will be posted on the workshop website.<br><br>*** Important dates ***<br><br>* Sunday, October 9th: Abstract submission deadline<br><br>* Sunday, October 23d: Notification of acceptance<br><br>* Friday, December 9th: Workshop<br><br><br>*** Contact ***<br><br><a href="mailto:main.at.nips2016@gmail.com" target="_blank">main.at.nips2016@gmail.com</a></div></div>
</div><br><br></div>