It's fascinating stuff going on there at Numenta.
Thanks for submitting this!<p>People have had huge expectations for so many decades in AI Research that it led to a bad image. But something with that is wrong, because every step towards understanding the brain brought all of us a step forward.<p>We know Nature is brilliant, so it's absolutely logical that it can't get copied by some funky scientists over the weekend. Could've been the great AI depression, but we'll get over it. I think the entire disappointment spiral led the AI Community to get treated like "wadda wadda wadda".<p>To me it appears like when a prehistoric human finds laser technology from a crashed UFO and doesn't know what to do with it. Making him disappointed about laser technology. So before whining about how slow we find things out about ourselves, the human brain and the mind we should better invest more in education. The more people can research this topic the faster we get amazing results.<p>When the research in a topic doesn't yield a promising result it doesn't mean the researchers are incompetent, it means we just don't have enough education or researchers who can complete the puzzle.<p>Before consuming popular information:
<a href="http://en.wikipedia.org/wiki/Harold_Lasswell" rel="nofollow">http://en.wikipedia.org/wiki/Harold_Lasswell</a><p>At the time the Numenta stuff got popular, the critics also got a voice: <a href="http://www.acceleratingfuture.com/michael/blog/2010/04/ben-goertzel-on-numenta/" rel="nofollow">http://www.acceleratingfuture.com/michael/blog/2010/04/ben-g...</a> and <a href="http://scienceblogs.com/pharyngula/2010/08/ray_kurzweil_does_not_understa.php" rel="nofollow">http://scienceblogs.com/pharyngula/2010/08/ray_kurzweil_does...</a><p>PS: enjoy the earworm: <a href="http://www.musick8.com/mclips/34wadda.mp3" rel="nofollow">http://www.musick8.com/mclips/34wadda.mp3</a>
Jeff Hawkins who also founded Palm is a cofounder of Numenta. He started the company after writing 'On Intelligence'. He does some great lectures that are on youtube.<p>This seems really cool, I signed up for the beta. I hope I am selected.
So some ML/AI researchers I know consider Numenta to be of a somewhat "outsider artist" effort. Their opinions of the research from Numenta is that of lower regard.<p>Now, you don't have to upvote me for this, but you don't have to downvote me, either, because I happen to know experts in ML/AI who disagree with Jeff Hawkins. Got that? Downvoting me doesn't change these other people's opinion. I'm just relating to you what I've been told by others more qualified and knowledgeable in an area than I am not an expert in. I have to say all this because people seem to give Numenta a very dogmatic approval and anyone who seems to question it on discussion boards gets flamed in a religious manner.<p>Edit: Hey, look, a downvote.<p>So, is Numenta genuinely really novel and new or is it snake oil? Or is it somewhere in between? Does it actually work better than what we have now?
How is it different from Google Prediction API? <a href="https://developers.google.com/prediction/" rel="nofollow">https://developers.google.com/prediction/</a>
It seems like HTMs would be prone to the same over-fitting problems that other types of neural networks have. I tried to find some comments in the Numenta material about this, but didn't see anything about Grok's strategy to avoid over-fitting. Can anyone help me figure this out?
If you're interested in learning more, Jeff Hawkins will be doing a keynote this year at Strange Loop talking about it in more depth. <a href="http://thestrangeloop.com/sessions" rel="nofollow">http://thestrangeloop.com/sessions</a>
On a related note what are some good algorithms for anomaly detection?<p>I want to run my credit card transactions through it to flag things for review.