Google TechTalks
1 HN recommender · recommended in 3 months, 2013–2017
1:02:06 · 84K views · Mar 11, 2013
Jeff Hawkins presents his view that the neocortex works on principles different from traditional computers. He describes sparse distributed representations and how they are learned, then shows how Numenta applies them to predict from high-velocity data streams.
Recommended as Hawkins' own accessible statement of his brain-inspired theory of intelligence and its software implementation.
3 comments from Hacker News
Here's an related talk by Jeff Hawkins at Google, where he gets into his ideas about the structure of the brain, and what he's done to reconstruct it in software/hardware. this video
I think that Hawkins has a great summary on some things that are necessary for intelligence [0]. Personally I think the tie between motor actuation and sensory detection is one of the most important -- did a project on that in 2010. Motor-sensory link is essentially how all intelligence evolved. Even c.elegans has neurons to determine whether it should wiggle one way or another. On the surface he seems to be cramming a whole lot of "solutions" into his HTM (hierarchical temporal memory). HTM is an interesting implementation and the sparse coding is definitely a benefit. However I think he is focused too much on his baby and not on other techniques that may fulfill the necessary components more efficiently. That is just on the surface. With a product / research balance maybe we just aren't seeing all the cool things going on underneath in research, but it does seem like that research will be shoe-horned into HTM whether or not it is the best architecture. [0] this video Starting at ~8m20s
What about this? Building Brains to Understand the World's Data (google tech talk): this video The result is [LINK]