Jeff Hawkins: Brains, Data, and Machine Intelligence [video] | Hacker News Reader

Jeff Hawkins: Brains, Data, and Machine Intelligence [video]

youtube.com

61 pointssuperfx15 comments

I was fortunate to discover Jeff Hawkins in the early 2007 , when he was exploring brain science .. .I went through his lectures for entrepreneurs at ecorner.stanford.edu , being indian ,I had to literally slow down the video to understand his speech and he is like Bill gross, very very fast in articulating and highly energetic in their talk. You can feel their passion in their work.

Incidently Jeff Hawkins was the first to work on Palm software.....you can say early versions of today's smartphones.

Coming back to brain science , I somehow feel it's more to do with complexity (Santafe.edu).

I am waiting for jeff Hawkins to launch a university like santafe.edu ( complexity) which works exclusively on brain science. Perhaps that would increase application of brain science.

I think this is the most interesting direction of research in CS one can get involved with today. In 10-20 years, half of CS graduate will be working in computational biology and the other half developing SDR-based machine intelligence.

It is not clear that these methods work better than "traditional" deep learning methods; in fact, they haven't produced any good results yet, and most experts think the hype around CLA is not because of its technical properties.

See e.g.

* http://www.reddit.com/r/MachineLearning/comments/25lnbt/ama_...

* http://www.reddit.com/r/MachineLearning/comments/2lmo0l/ama_...

* http://www.reddit.com/r/MachineLearning/comments/2iejpg/syst...

Can "traditional deep learning methods" replicate the kind of unsupervised learning demonstrated by cortical.io's semantic retina (fox eats rodent example in the video)?

Sure they can, see e.g. http://nlp.stanford.edu/pubs/SocherChenManningNg_NIPS2013.pd... or http://arxiv.org/pdf/1301.3618v2.pdf.

Check the first paper - when working on this problem researchers from Stanford developed a way to measure the quality of an approach, evaluated the results on 30k+ relation examples from 2 different datasets and compared their algorithm with 4 other algorithms.

The problem with cortical.io or numenta (both commercial companies) is that they don't compare their approaches with existing approaches and don't evaluate them on public datasets. And when people do such comparison existing approaches turn out to be better.

It is totally possible these algorithms are good and they provide something that "traditional" methods don't provide. But this is yet to be shown; for some reason authors decide not to "compete" on a same ground. Instead of "promoting" their methods in scientific community via publications / comparisions with existing approaches they seem to focus on people who have little knowledge of modern machine learning. Also, they use their own terminology and usually refer only to their own papers or to some obscure papers from ten years ago, which doesn't help.

>"But this is yet to be shown; for some reason authors decide not to "compete" on a same ground. Instead of "promoting" their methods in scientific community via publications / comparisions with existing approaches they seem to focus on people who have little knowledge of modern machine learning."

You'd have to provide them then with a decent argument about what economic/competitive benefit they would get by what you suggest. You say they're commercial companies, so then don't be surprised when their approach is based on financial incentives. But trust me, they're probably begging to have someone show them a better alternative that will give them a competitive edge.

So, either no one like you has given them that alternative/idea. Someone has already, and they rejected the financial benefit. Or, finally, someone already told them your idea but they discovered there was no financial benefit and the only benefit was for the greater society.

A cynical view on this could be the following: Numenta sells licenses, cortical.io sells api requests, they benefit from more developers using their tools. Commercial companies which fund deep learning research (like Microsoft and Google ) develop their own products which are based on machine learning, they benefit from advancing state of the arts, from better algorithms. So we have quality publications and algorithms from Microsoft or Google and quality marketing from Numenta or cortical.io.

The idea of representing words as vectors and testing was originally a deep learning thing. See Google's word2vec which is famous for being able to do things like "king"-"man"+"woman"= "queen".

Even before Googles word2vec at Berkeley Lab they were experimenting with this kind of vector space 'a search engine that thinks, almost' http://newscenter.lbl.gov/2005/03/31/a-search-engine-that-th...

As far as this technology will deliver I belive it will become a engineering or scientific branch in its own right. I suspect it will be a bit like economics, with lots of clever people deriving sophisticated models but expert knowledge is all about heuristics (at least this is my layman view of economics): anyone can understand the basic principles but one would be hard pushed to claim there is anyone who really grasps the thing in entirety.

... except those who control it. :-)

What might be interesting is to stream stock market data (trades as well as company's press releases) into the Grok algorithm. People who will manage this by the next market crash will be very rich indeed.

Im working on this now, but not using grok, using NuPIC core directly. Its been fun maintaining my own private fork of the repo while I scale the algorithms to run on > 1 machine i.e. be able to build a model of any topology where the regions aren't necessarily on the same machine, but still must be fast.

Please stop talking about machine intelligence without reference to normatively correct principles of quantitative reasoning. "I tried it and it scored well in validation" just shows that you've managed to apply some kind of statistical principle, not that you've found a way to build complete machine intellect.

Does Hawkins do that? I've not watched this one yet, but his talks are always much more high-level than that I think, and while his work is under the general umbrella of artificial intelligence, what he is actually trying to do is to emulate the human neo-cortex in gritty detail.

Reply on news.ycombinator.com