IBM's Lectures on Cognitive Computing

IBM's Lectures on Cognitive Computing

• Lecture 1: From Brain Dynamics to Consciousness by Gerald Edelman (The Neurosciences Institute)
  Subtitle: A Prelude to the Future of Brain-Based Devices
  Slides of Lecture 1
  
  Edelman discusses neuronal group selection, brain-based devices, and robots playing soccer.



• Lecture 2: The Emergence of Intelligence in the Neocortical Microcircuit by Henry Markram (EPFL/BlueBrain)
  Slides of Lecture 2
  
  Markram discusses microcolumns in the brain, and shows several video animations of computer models of neurons communicating in a microcolumn. His model includes 10,000 neurons, which is a very large number of neurons to model.



• Lecture 3: The Mechanism of Thought by Robert Hecht-Nielsen (UCSD)
  Slides of Lecture 3
  
  Hecht-Nielsen discusses sentence generation based on knowledge links. It's so good, it appears to generate new sentences based on strong semantic understanding of the input sentences. He also demonstrates robust speech understanding. His work is sponsored by Fair Isaac and the Office of Naval Research.



• Lecture 4: Hierarchical Temporal Memory: Theory and Implementation by Jeff Hawkins (Palm/Numenta)
  Subtitle: A new computational paradigm based on cortical theory
  Slides of Lecture 4



• Lecture 5: How the brain works, what it computes, and how/when we might build one
  Part I: A Scientific Theory of Mind by James Albus (NIST)
  Goal of his research is to extend the frontiers of human knowledge to include a scientific understanding of the processes in the human brain that give rise to the phenomenon of mind.
  Slides of Part I
  Part II: Theodore Berger (USC): Discusses how to replace parts of the brain with VLSI computer chips (in the case of bringing functionality back to a region of the brain that was removed due to a tumor).
  Slides of Part II
  Part III: Kwabena Boahen (Stanford): Presents energy analysis of brains vs. computers.
  Slides of Part III
  Part IV: Ralph Linsker (IBM): Shows demo of separating multiple overlapping voices using a neurally-inspired algorithm.
  Slides of Part IV
  Part V: Jerry Swartz (The Swartz Foundation): Presents Large scale brain modeling.
  Slides of Part V



• Lecture 6: The Uniqueness of the Human Brain by V. S. Ramachandran (UCSD)
  Slides of Lecture 6
  
  Ramachandran discusses phantom limbs and synesthesia (esp. color-blind synesthete) as a function of neuron innervation. Specifically, he focuses on cross-linking between nearby cortical regions, which he believes to be genetically caused (e.g. synesthesia appears to be found frequently in family lineages). He also discusses the link between mirror neurons and autism, and how language invention is due to an inherent cross-linking between portions of the visual and auditory regions (e.g. Buba/Kiki effect).



• Lecture 7: Beyond Dualism by John Searle (UC Berkeley)
  (no slides available)



• Lecture 8: Cortical Dynamics of Working Memory by Joaquin Fuster (UCLA)
  Slides of Lecture 8



• Lecture 9: A Quantitative Theory of Cortex by Leslie Valiant (Harvard University)
  (no slides available)



• Lecture 10: The 4 C's of Neuroinformation Theory by Toby Berger (University of Virginia)
  Subtitle: Coding, Computing, Control and Cognition
  Slides of Lecture 10



• Lecture 11: Consciousness by Christof Koch (Caltech)
  (no slides available)



• Lecture 12: The Future of Cognitive Computing by William Pulleyblank (IBM Global Services)
  (no slides available)