AI DEBATE 3
The Debate The AI World Is Waiting For
MONTREAL.AI
Fri, Dec 23, 2022 6:00 PM - 8:30 PM EST
Moderator and co-organizer (with Vincent Boucher): Gary Marcus
Confirmed speakers: "TBA"
Schedule
Panel 1: "TBA"
Panelists TBA
Panel 2: "TBA"
Panelists TBA
Panel 3: "TBA"
Panelists TBA
AI DEBATE 2
Moving AI Forward: An Interdisciplinary Approach
MONTREAL.AI
Wed, December 23, 2020 | 4:00 PM - 7:00 PM EST
Moderator and co-organizer (with Vincent Boucher): Gary Marcus
Confirmed speakers: Ryan Calo, Yejin Choi, Daniel Kahneman, Celeste Kidd, Christof Koch, Luis Lamb, Fei-Fei Li, Adam Marblestone, Margaret Mitchell, Robert Osazuwa Ness, Judea Pearl, Francesca Rossi, Ken Stanley, Rich Sutton, Doris Tsao, Barbara Tversky and more TBA
Schedule
Panel 1: "Architecture and Challenges"
Yejin Choi, Luis Lamb, Fei-Fei Li, Robert Ness, Judea Pearl, Ken Stanley and Rich Sutton
Panel 2: "Insights from Neuroscience and Psychology"
Danny Kahneman, Christof Koch, Adam Marblestone, Doris Tsao and Barbara Tversky
Panel 3: "Towards AI we Can Trust"
Ryan Calo, Celeste Kidd, Margaret Mitchell and Francesca Rossi
Final Words, Readings, Slides and Relevant Materials
"It takes a village to raise an AI that's ethical, robust, and trustworthy" - Gary Marcus
Ryan Calo
Remark
*Ryan Calo's Remark at AI DEBATE 2: Artificial Intelligence Policy: Not Just A Matter of Principles, Ryan Calo, 2020: https://youtu.be/XoYYpLIoxf0
Yejin Choi
Remark
*Yejin Choi's Remark at AI DEBATE 2: Commonsense AI: Cracking the Longstanding Challenge in AI, Yejin Choi, 2020: https://youtu.be/fs80dg6hE3U
Daniel Kahneman
Remark
*Daniel Kahneman's Remark at AI DEBATE 2: System 1 is not non-symbolic, Daniel Kahneman, 2020: https://youtu.be/2zNd69ZGZ8o
Celeste Kidd
Remark
*Celeste Kidd's Remark at AI DEBATE 2: Profound impacts of AI (and its biases) on human beliefs, Celeste Kidd, 2020: https://youtu.be/chrGgpTVH2g
Christof Koch
Remark
*Christof Koch's Remark at AI DEBATE 2: Don’t look (anymore) at neuroscience for help with AI, Christof Koch, 2020: https://youtu.be/2KqmX4l4QYc
Luis Lamb
Remark
*Luis Lamb's Remark at AI DEBATE 2: Neurosymbolic AI: The 3rd Wave, Luis Lamb, 2020: https://youtu.be/sQtDoAEvU3Q
Final Words
"The AI Debate #2 hosted a convergence of diverse views on the future of AI.
We need from a technical perspective to integrate machine learning and logical) reasoning to build interpretable and explainable AI systems and technologies. This is in line with systems 1 and 2 of Kahneman: we need to integrate AI schools of thought, which ideally walk hand in hand and thus strengthen science.
We have to improve scientific understanding amongst AI paradigms, so as to build AI that benefits humanity and the planet. The world will need a principled AI education for all, since AI will be the key technology of the next decades, if not of the XXI century." - Luis Lamb
Readings and Relevant Materials
"Readings that harmonize with the great AI Debate #2" - Luis Lamb
Neurosymbolic AI: The 3rd Wave, Artur d’Avila Garcez and Luis Lamb, 2020: https://arxiv.org/abs/2012.05876
Thinking Fast and Slow in AI, Booch et al., 2020: https://arxiv.org/abs/2010.06002
Neural-Symbolic Cognitive Reasoning, D'Avila Garcez, Artur S., Lamb, Luís C., Gabbay, Dov, 2009: https://www.springer.com/gp/book/9783540732457
Rebooting AI : Building Artificial Intelligence We Can Trust, Gary Marcus and Ernest Davis, 2019: http://rebooting.ai
Causality, Judea Pearl, 2009 (2nd Edition): http://bayes.cs.ucla.edu/BOOK-2K/
Probably Approximately Correct: Nature's Algorithms for Learning and Prospering in a Complex World, Leslie Valiant, 2013: https://dl.acm.org/doi/book/10.5555/2536711
Learning to Solve NP-Complete Problems: A Graph Neural Network for Decision TSP, Prates et al., 2019: https://ojs.aaai.org//index.php/AAAI/article/view/4399
Learning a SAT Solver from Single-Bit Supervision, Selsam et al., 2019: https://openreview.net/forum?id=HJMC_iA5tm
Graph Neural Networks Meet Neural-Symbolic Computing: A Survey and Perspective, Lamb et al., 2020: https://www.ijcai.org/Proceedings/2020/679
Thinking, Fast and Slow, Daniel Kahneman, 2011: https://en.wikipedia.org/wiki/Thinking,_Fast_and_Slow
Fei-Fei Li
Remark
*Fei-Fei Li's Remark at AI DEBATE 2: In search of the next AI North Star: a tale of two kittens, Fei-Fei Li, 2020: https://youtu.be/XY1VTLRIsNo
Adam Marblestone
Remark
*Adam Marblestone's Remark at AI DEBATE 2: Leveraging neuro-technology for AI, Adam Marblestone, 2020: https://youtu.be/gLVmky_BD5c
Margaret Mitchell
Remark
*Margaret Mitchell's Remark at AI DEBATE 2: Ethics in the Vision and Language of Artificial Intelligence, Margaret Mitchell, 2020: https://youtu.be/i8LlrhsQ-lQ
Robert Ness
Remark
*Robert Ness's Remark at AI DEBATE 2: Causal Reasoning with (Deep) Probabilistic Programming, Robert Ness, 2020: https://youtu.be/qHOfSoc4Oqc
Judea Pearl
Remark
*Judea Pearl’s Remark at AI DEBATE 2: The Domestication of Causal Reasoning: Cultural and Methodological Implications, Judea Pearl, 2020: https://youtu.be/gJW3nOQ4SEA
What I would have said had I been given six (6), instead of three (3) minutes, Judea Pearl, 2020: http://causality.cs.ucla.edu/blog/index.php/2020/12/28/edited-script-of-j-pearl-talk-at-montreal-ai-debate-2/
Readings and Relevant Materials
The Seven Tools of Causal Inference, Judea Pearl, 2016: https://ucla.in/2HI2yyx (Summary: https://vimeo.com/314324108)
Radical Empiricism and Machine Learning Research, Judea Pearl, 2020: https://ucla.in/32YKcWy
Data versus Science: Contesting the Soul of Data-Scienc, Judea Pearl, 2020: https://ucla.in/3iEDRVo
Francesca Rossi
Remark
*Francesca Rossi’s Remark at AI DEBATE 2: Thinking Fast and Slow in AI: Towards more general and trustworthy AI, Francesca Rossi, 2020: https://youtu.be/XE-WThtchuM
Kenneth O. Stanley
Slides
*Kenneth O. Stanley’s Slides presented at AI DEBATE 2: Open-Endedness, Evolution, and AI, Kenneth O. Stanley, 2020: https://montrealartificialintelligence.com/aidebate2/slidesstanley.pdf
Reading
Why Greatness Cannot Be Planned: The Myth of the Objective, Kenneth O. Stanley and Joel Lehman, 2015th Edition: https://www.amazon.com/Why-Greatness-Cannot-Planned-Objective/dp/3319155237
Rich Sutton
Remark
*Rich Sutton's Remark at AI DEBATE 2: Reinforcement Learning is the Computational Theory of Intelligence, Rich Sutton, 2020: https://youtu.be/hcJNFdZit-Q
Doris Tsao
Remark
*Doris Tsao's Remark at AI DEBATE 2: How the brain builds a model of the world: insights from neuro-science, Doris Tsao, 2020: https://youtu.be/FkKUAGSNHlE
Barbara Tversky
Remark
*Barbara Tversky's Remark at AI DEBATE 2: Thinking with the Body and the World, Barbara Tversky, 2020: https://youtu.be/OFRY39ansJc
Press
Leading computer scientists debate the next steps for AI in 2021, By Ben Dickson | January 2, 2021 | VentureBeat: https://venturebeat.com/2021/01/02/leading-computer-scientists-debate-the-next-steps-for-ai-in-2021/
AI Debate 2: Night of a thousand AI scholars, By Tiernan Ray | December 23, 2020 | ZDNet: https://www.zdnet.com/article/ai-debate4-2-night-of-a-thousand-ai-scholars//
‘The Debate of the Next Decade’ – AI Debate 2 Explores AGI and AI Ethics, Reporter: Yuan Yuan | Editor: Michael Sarazen | December 24, 2020 | Synced: https://syncedreview.com/2020/12/24/the-debate-of-the-next-decade-ai-debate-2-explores-agi-and-ai-ethics/
Pre-Readings
Pre-Readings https://montrealartificialintelligence.com/aidebate2/readings.pdf
DEBATE : Yoshua Bengio | Gary Marcus
The Best Way Forward For AI
MONTREAL.AI
Monday, 23 December 2019 from 6:30 PM to 8:30 PM (EST) at Mila
REPORT ON THE AI DEBATE : https://medium.com/@Montreal.AI/report-on-the-ai-debate-c62fcbf2ca43
Yoshua Bengio and Gary Marcus on the best way forward for AI
Moderated by Vincent Boucher
Gary Marcus, slides : https://montrealartificialintelligence.com/aidebate/slidesmarcus.pdf
Yoshua Bengio, slides : https://montrealartificialintelligence.com/aidebate/slidesbengio.pdf
Transcript of the AI Debate : https://medium.com/@Montreal.AI/transcript-of-the-ai-debate-1e098eeb8465 | PDF
After AI Debate Discussion
Bengio-Marcus AI Debate Post Mortem, Part I: The Deep Learning Pivot, Gary Marcus, Dec. 26, 2019 : https://medium.com/@GaryMarcus/bengio-marcus-ai-debate-post-mortem-part-i-the-deep-learning-pivot-f7bd62b9861c
Response to Gary Marcus, Yoshua Bengio, December 26th 2019 : https://docs.google.com/document/d/1P9YyZ4xEDO98qPTa1Al4RnTqxj3mMiCvQE4i3hZkthY/edit
A research program is not a set of techniques: A brief response to Yoshua Bengio’s December 26 reply to me, Gary Marcus, Dec. 27, 2019 : https://medium.com/@GaryMarcus/a-research-program-is-not-a-set-of-techniques-a-brief-response-to-yoshua-bengios-december-26-fafc0a29ffc9
Yoshua's definition of Deep Learning, Yoshua Bengio, December 27th 2019 : https://github.com/MontrealAI/MontrealAI.github.io/blob/master/aidebate/yoshuadefinitiondeeplearning.png
Deep learning, science, engineering, research, and terminology, Gary Marcus, January 1, 2020 : https://medium.com/@GaryMarcus/deep-learning-science-engineering-research-and-terminology-292a747a94d3
The Next Decade in AI: Four Steps Towards Robust Artificial Intelligence, Gary Marcus, Feb 14, 2020 : https://arxiv.org/abs/2002.06177
Readings by Yoshua & Gary
Readings by Yoshua
From System 1 Deep Learning to System 2 Deep Learning, Yoshua Bengio, NeurIPS, Dec. 11, 2019 : https://slideslive.com/38921750/from-system-1-deep-learning-to-system-2-deep-learning
The Consciousness Prior, Bengio et al., 2017 (last revised 2 Dec 2019) : https://arxiv.org/abs/1709.08568
Recurrent Independent Mechanisms, Goyal et al., 2019 : https://arxiv.org/abs/1909.10893
Learning Neural Causal Models from Unknown Interventions, Ke et al., 2019 : https://arxiv.org/abs/1910
BabyAI: First Steps Towards Grounded Language Learning With a Human In the Loop, Chevalier-Boisvert et al., 2018 (latest version 19 Dec 2019) : https://arxiv.org/abs/1810.08272v4
A Meta-Transfer Objective for Learning to Disentangle Causal Mechanisms, Bengio et al., 2019 : https://arxiv.org/abs/1901.10912
Readings by Gary
Rule learning by seven-month-old infants, G. F. Marcus et al., Science 283(5398):77-80, February 1999 : https://www.researchgate.net/publication/13415195_Rule_learning_by_seven-month-old_infants
Rebooting AI : Building Artificial Intelligence We Can Trust, Gary Marcus and Ernest Davis, 2019 (Chapter 4, for Bengio and Marcus debate) : https://cs.nyu.edu/faculty/davise/Rebooting/Chapter4.pdf
The Algebraic Mind, Gary F. Marcus, 2001 (chapters 2 (last part) and 3, now available free from MIT Press, in celebration of the debate) : https://mitpress.mit.edu/books/algebraic-mind
Innateness, AlphaZero, and Artificial Intelligence, Gary Marcus, 2018 : https://arxiv.org/abs/1801.05667
The Birth of the Mind (Chapters 6 - 8), Gary Marcus, 2004
Rethinking Eliminative Connectionism, Gary Marcus, 1998 : https://www.sciencedirect.com/science/article/pii/S0010028598906946
Deep Learning: A Critical Appraisal, Gary Marcus, 2018 : https://arxiv.org/abs/1801.00631
Pre-Readings
Pre-Readings by Gary & Yoshua : http://www.montreal.ai/aidebate.pdf
Press
Devil's in the details in Historic AI debate, By Tiernan Ray | December 24, 2019 | ZDNet : https://www.zdnet.com/article/devils-in-the-details-in-bengio-marcus-ai-debate/
Debate do ano em IA cita trabalho de Lamb, Jornal do Comercio | December 27, 2019 | ZDNet : https://www.jornaldocomercio.com/_conteudo/colunas/mercado_digital/2019/12/718693-debate-do-ano-em-ia-cita-trabalho-de-lamb.html
What's in a name? The 'deep learning' debate, By Tiernan Ray | December 28, 2019 | ZDNet : https://www.zdnet.com/article/whats-in-a-name-the-deep-learning-debate/
Do you need a prior? What about hybrid models? Marcus and Bengio debate the future of AI, 機器之心報道, 參與：一鳴、張倩、蛋醬 | December 25, 2019 | ChainNews : https://www.chainnews.com/zh-hant/articles/321502060639.htm
[Big Boss] Yoshua Bengio and Gary Marcus' Christmas AI Debate "The Best Way Forward for AI" with video and slides, By Quan Zhuanzhi | December 25, 2019 : https://wemp.app/posts/313a2a2e-e014-4c22-bfea-9b3e3bbbdf77
MONTREAL.AI Debates Series