Archived VideoPanels

Sessions below are arranged in the reverse chronological order. Yet, a better layout is available:


[ view all archived VideoPanels here ]


VP8: Paper discussion with Stephen Kosslyn

(detailed info will be posted shortly)


VP7: Discussion of the paper by C. Eliasmith et al. (2012) Science 338: 1202-1205

  • Date and Time: March 8th, 2013 (Friday), started at 2:35 PM Eastern Time, duration 28 min
  • Panelists: Chris Eliasmith (Philosophy, SDE, CRCTN, and CTN, U Waterloo) and Terry Stewart (CNRG, CTN, U Waterloo)
  • Leaders-Moderators: Mahan Mollajafar (Senior, Neuroscience, College of Humanities & Social Sciences, George Mason U) and Omead Freshtvadi (Junior, Psychology, College of Humanities & Social Sciences, George Mason U)
  • Host: Alexei Samsonovich
  • Topic: Spaun Paper Discussion
  • Title: A large-scale model of the functioning brain
  • Authors: Chris Eliasmith, Terrence C. Stewart, Xuan Choo, Trevor Bekolay, Travis DeWolf, Yichuan Tang, and Daniel Rasmussen
  • Published in: Science 338: 1202-1205 (2012)
  • Abstract: A central challenge for cognitive and systems neuroscience is to relate the incredibly complex behavior of animals to the equally complex activity of their brains. Recently described, large-scale neural models have not bridged this gap between neural activity and biological function. In this work, we present a 2.5-million-neuron model of the brain (called “Spaun”) that bridges this gap by exhibiting many different behaviors. The model is presented only with visual image sequences, and it draws all of its responses with a physically modeled arm. Although simplified, the model captures many aspects of neuroanatomy, neurophysiology, and psychological behavior, which we demonstrate via eight diverse tasks.
  • Discussion Agenda: To address questions listed below that were prepared by the moderators.
  1. Has there been any recent result or progress made based on your project?
  2. What do you think is the biggest scientific contribution of your work?
  3. Can your model still work if changes were made to settings and environment?
  4. Do you believe your model is robust or flexible enough to handle a different paradigm without adjustment (e.g. processing two-digit numbers)? If not, then how do you rank its performance compared to human performance?
  5. Do you believe that attention control can allow your model to succeed in circumstances when an unexpected stimulus (e.g., laser pointer) is introduced during the task, or the target stimulus is moved to a side of the visual field?
  6. Why do you believe that it is important to have spiking neurons or, more generally, neurons as the basis elements of your model, while your paradigm is set to perform a certain task or produce a certain behavior?
  7. In your paper, you were able to kill neurons and show robustness of the model performance. Can your model also display robustness with respect to neurogenesis? Can it make use of newly generated neurons?
  8. Do you believe your model is an essential step in science or a sidetrack?
  9. Is your model sufficiently accurate to generate valid predictions, as a model of the human brain? Or to display realistic adaptation to unexpected conditions? Will it reveal unexpected new information about the brain?
  10. How useful or effective can your model be in simulations of neurological disorders and their treatment (e.g. Parkinson, Alzheimer’s)?
  11. Your model only includes 2.5 million neuron, while the actual brain contains about 100 billion neurons. With the small fraction of neurons represent in your model, how valid can be its simulation?
  12. With respect to other projects, like the Blue Brain project of IBM or the Cognitive Computation project, both representing a greater scale of the number of neurons compared to Spaun, what makes your model better and more interesting?
  13. Can your model reproduce emotions?

Or, watch it on Vimeo: http://player.vimeo.com/video/61374659





VP6: Panel Discussion of the paper on LIDA & GWT


  • Date and Time: August 27th, 2012 (Monday), started at 12:05 PM Eastern Time, duration 13 min
  • Panelists: Stan Franklin (Professor of Computer Science, University of Memphis) and Bernard Baars (Affiliated Research Fellow, Krasnow Institute for Advanced Study, George Mason University)
  • Host-Moderator: Alexei Samsonovich
  • Title: Global Workspace Theory, its LIDA model and the underlying neuroscience
  • Authors: Stan Franklin, Steve Strain, Javier Snaider, Ryan McCall, Usef Faghihi
  • Published in: Biologically Inspired Cognitive Architectures 1: 32-43 (2012).
  • Abstract: A biologically inspired cognitive architecture must draw its insights from what is known from animal (including human) cognition. Such architectures should faithfully model the high-level modules and processes of cognitive neuroscience. Also, biologically inspired cognitive architectures are expected to contribute to the BICA ‘‘challenge of creating a real-life computational equivalent of the human mind’’. One unified theory of cognition, Global Workspace Theory (GWT) has emerged as the most widely accepted, empirically supported theory of the role of consciousness in cognition. Recent experimental studies reveal rich cortical connectivity capable of supporting a large-scale dynamic network. We propose that brains in fact cyclically and dynamically form such a network according to GWT…
  • Discussion Agenda: This Videopanel is linked to the above journal publication and complements the paper. Guiding questions are:
  1. In what sense can we say that LIDA implements the Global Workspace Theory?
  2. In LIDA, the global workspace theory is used for action selection. Why not use it for sensory perception, language understanding and production, problem solving, etc.?
  3. Does the global workspace theory really explain consciousness?
  4. Is there any connection between the functional consciousness implemented in LIDA and the phenomenal consciousness?

Or, watch it on YouTube: http://www.youtube.com/embed/WilYeIfsthQ?rel=0





VP5: Panel Discussion on Emotional AI, Part B

  • Date and Time: February 9th, 2012 (Wednesday), started at 4:10 PM Eastern Time, duration 46 min
  • Panelists: Eva Hudlicka (Psychometrix), Charles Peck (IBM Watson), Mike Sellers (Online Alchemy), Rodrigo Ventura (http://users.isr.ist.utl.pt/~yoda)
  • Leader & Moderator: Mike Sellers
  • Audience: Christine Lisetti, Ricardo Sanz, Rosario Sorbello
  • Host: Alexei Samsonovich
  • Topic: Emotional AI
  • Agenda: This Videopanel is a continuation of the discussion that started a day ago and is available below. Both sessions are intended to help with writing a seminal vision paper with a tentative title "The emergence of emotional artificial intelligence" for the inaugural issue of Elsevier's BICA. The agenda in this session was to address questions listed below.
  1. How the following notions differ from each other, and how do they interact? Emotion, affect, appraisal, arousal, motivation, feeling, mood, desire, preference, subjective experience, quale (qualia)...
  2. What can you say about the taxonomy of emotions (assuming that we can use the word)?
  3. How do emotions of different agents affect each other in a social context?
  4. Why should computer scientists study emotions?
  5. In what sense should AI be emotionally intelligent? And emotional itself?
  6. What is your vision for the next 50 years of emotional AI?
View clips separately with better sound: [ Clip 1] , [ Clip 2 ]; [ download audio: vp20120209.mp3, 90mb ]





VP4: Panel Discussion on Emotional AI, Part A

  • Date and Time: February 8th, 2012 (Wednesday), started at 10:15 Eastern Time, duration 20 min
  • Panelist: Rony Novianto (IBM PhD Fellowship Awards Holder, Australian Endeavour Awards Holder, www.ronynovianto.com)
  • Host & Moderator: Alexei Samsonovich
  • Topic: Emotional AI
  • Agenda: This Videopanel is intended to help with writing a seminal vision paper with a tentative title "The emergence of emotional artificial intelligence" for the inaugural issue of Elsevier's BICA. There were a number of talks and discussions on this topic at BICA 2011. Therefore, it seems sensible to invite those authors and presenters to write a collective paper. Given that the topic is highly controversial, after some initial discussions it was decided to divide the paper into sections - one for each co-author or group - representing particular views, theories and schools in terms of answers to one and the same list of questions. The introduction and conclusions will be collectively edited by all. Therefore, the agenda for this multi-part Videopanel includes 3 main themes:
  1. What schools of thought, theories or individual views on emotions in AI will be represented in each section, and by whom?
  2. What common set of questions should be addressed by each two-page section of the 18-page paper? A candidate list is given below.
  3. What are your brief preliminary ideas of answers to those questions, and (if the time allows) your criticism of others? Can we come to any constructive conclusion - or should we just conclude that we all mutually disagree?
  • Questions:
  1. How the following notions differ from each other, and how do they interact? Emotion, affect, appraisal, arousal, motivation, feeling, mood, desire, preference, subjective experience, quale (qualia), ...
  2. What is the taxonomy of emotions (assuming that we can use the word)?
  3. How do emotions of different agents affect each other in a social context?
  4. Why should computer scientists study emotions?
  5. What VR paradigms can be used for modeling studies of emotional interactions in social groups?
  6. In what sense should AI be emotionally intelligent?
  7. What approach should be taken for implementation of emotional intelligence in AI?
  8. In the last two questions, what about higher or complex emotions? (examples: pride, shame, resentment, compassion, humor)
[ download audio: vp20120208.mp3, 38mb ]





VP3: Panel Discussion on Metacognition

  • Date and Time: June 19th, 2011 (Sunday), started at 2:30 Eastern Time, duration 1 hour 9 minutes
  • Panelists: Aaron Sloman, Ashok Goel, Michael Anderson, Simon Levy
  • Audience: Scott E. Fahlman, others unidentified...
  • Host & Moderator: Kamilla Jóhannsdóttir
  • Topic: Metacognition
  • Agenda: The 3 guiding questions are:
  • (1) People talk a lot about metacognition in AI, and yet there is little conclusive proof of its leverage in cognition and learning. How and why should cognitive architectures benefit from metacognition?
  • (2) What is metacognition in AI? Perhaps the biggest controversy is in understanding of this concept by different researchers. What separates metacognitive architectures from merely cognitive architectures? Also, how is metacognition related to creativity, self-regulation, self-awareness, theory of mind, higher-order thoughts, complex emotions, social cognition, and finally, the ability to learn like a human?
  • (3) Is it possible to have a universal metacognitive assistant that can improve performance of virtually any merely cognitive system connected to it? Or, is metacognition and its implementation always domain-specific?
  • TRANSCRIPT AVAILABLE


Download this video (560 Mb)




VP2: Paper Presentations

  • Date and Time: June 17th, 2011 (Friday), started at 12:00 Eastern Time, duration 45 minutes
  • Presenters: Vladimir Redko (Deputy Director for Research, Center of Optical Neural Technologies, Scientific Research Institute for System Analysis, Russian Academy of Sciences) and Evgenii Vityaev (Professor, Leading Scientist, Sobolev Institute of Mathematics of the Siberian Branch of the Russian Academy of Sciences)
  • Panelists: Andrew Nuxoll, Simon Levy, Brandon Rohrer, Antonio Chella
  • Audience: (Richard) Chi Leung Lau, Elena Fedorovskaya, David Games, Don Sofge, Eugene Surowitz, Tomas Singliar, Rony Novianto, Meehae Song, Charles Peck, Miguel Salichs...
  • Host & Moderator: Kamilla Jóhannsdóttir
  • Topic, Agenda and Materials: Presentation of Paper #26 (Vladimir Redko and Anton Koval, Evolutionary Approach to Investigations of Cognitive Systems) and Paper #50 (Evgenii Vityaev and Alexander Demin, Animat –°ontrol System based on Functional Systems Theory). Slides are available (RedkoKoval.ppt, Vityaev_Demin_2003.ppt, Vityaev_Demin_2003.pdf).







VP1: Panel Discussion on the BICA Challenge

  • Date and Time: May 6th, 2011 (Friday), started at 1:00 pm US Eastern Time, duration 0:18:34
  • Panelists: Andrea Stocco (U Washington), Antonio Chella (U Palermo), and Brandon Rohrer (Sandia)
  • Host & Moderator: Alexei Samsonovich
  • Topic and Agenda: Panelists in turn express their views on the BICA Challenge guided by a list of questions
  • Supplementary Materials: Slides are available (bicavp20110506.ppt)
  • Remarks: Antonio Chella disappeared in the middle of the session because he experienced a crash in the Department network.
  • Technical Tips: You may need to expand your browser to full screen. If the video still does not fit, press the Full Screen button in the video (bottom right corner, after HD button). Also, you should use the Full Screen mode if your screen is exactly 1280 pixels wide.
If you cannot play the movie below in your browser, you can try:
- download the movie: bicavp20110506.mov.zip
- download bicavp20110506.mov.zip from https://public.me.com/samsonovich (get passwd)
- play it directly from Vimeo: http://vimeo.com/23378186
- play it from YouTube: http://www.youtube.com/watch?v=7apAYP4_M6Q






Advanced Web Counter