readwriteweb.com — Jeff Hawkins invented the Palm Pilot and wrote a book called 'On Intelligence'. Now he has founded Numenta - a company dedicated to developing algorithms and software based on the ideas put forward in the book. This spring Numenta released its first product, an experimental software aimed at researchers and advanced developers.
Mar 27, 2007 View in Crawl 4
bobfosterMar 27, 2007
Looks like snake oil, smells like snake oil, it must be...AI.
databyssMar 27, 2007
You are a little late my friend:"Skynet 5 sent into orbitA SATELLITE that will provide Britain with the latest in military communications technology has been successfully deployed in space."<a class="user" href="http://news.scotsman.com/uk.cfm?id=392522007">http://news.scotsman.com/uk.cfm?id=392522007</a>
darkstar949Mar 27, 2007
The way humans think is complex - but in essence you are correct and most AI research is not so much focuses on the goal of creating an general intelligence, but rather creating an intelligent means of addressing a specific problem.
bodhidharmazenMar 27, 2007
dreadsword,First of all, I do not believe at all the "common sense" about the brain being "a computer". That said, I have difficult also in consider consciousness as an "emergent property of brain activity". At best, thats an ad hoc hypothesis, but no more.Now, regarding structures, I do believe that what we call consciousness is based on the structure of the brain and of matter, in a general sense. So, I do not believe that, even in principle, it would be impossible to "create" a form of AC. While it wont be based on current technology, I do believe a sentient machine is viable.
sotopheavyMar 28, 2007
This model of human cognition is one of hundreds. The problem with things like this is that every time somebody thinks of a new way the brain might work they have to start from scratch and create a model. What would really help is a set of tools that could speed up this process. There are a whole lot more ideas out there than can be implemented. I do believe that AI will develop not only in small sense, but in every sense there is. As AI technology develops further it will also aide researchers in developing new and better models quicker... hopefully sans terminator/Hal.
cbrackMar 28, 2007
@logicbomb...I being your father, choose to teach you from birth how to add numbers, get my news, organize and retrieve my filed papers, and hold up a little "start" (thats what I'd call it) menu that I can choose commands for you to do. I'll also have you throw up a blue piece of paper with white text describing any errors you encounter while doing those tasks.Now, I must ask you, how would that make you much different from windows 98?We may not be able to create a 99% human robot, but we can surely build a robot that can learn and seem just as human as anyone.
flankerMar 28, 2007
I'm going to have to disagree with you there. See my comment above under fr0z3nph03n1x.
flankerMar 28, 2007
@exodii: wow you bought into Hawkins' spin hook, line, and sinker didn't you? On Intelligence is an incredibly arrogant treatise in which Hawkins insults pretty much the entire field of neurophysiology. There's a reason why even people who have spent their entire lives studying the cortex refuse to advance a single theory as to "how it all works." It's too complicated for that. Hawkins focuses on a single paper, dismisses everything else as a minor detail and bases his theory on that premise. I'm not saying it's not going to produce some interesting computational results, but to herald it as THE explanation of how the biological cortex works is premature and almost certainly incorrect.
flankerMar 28, 2007
@williamdyer: That's an overly simplistic view which fails to explain a lot of observations in neurophysiology.
flankerApr 6, 2007
@williamdyer: I can't tell if you're trolling or not, but if not, you clearly have no idea what you're talking about.Here's a stain of the visual cortex: <a class="user" href="http://hubel.med.harvard.edu/97.jpg">http://hubel.med.harvard.edu/97.jpg</a>Here's a stain of frontal cotex (use the control column): <a class="user" href="http://www.nature.com/emboj/journal/v17/n3/images/7590788f2.jpg">http://www.nature.com/emboj/journal/v17/n3/images/7590788f2.jpg</a>Do those look the same to you?As for your firm belief that regions of the brain can "reprogram" themselves -- this is true only to a minor degree and not to the degree you'd expect if everything was exactly the same and modular and hot-swappable as you make it seem. If that were true, then I see no reason for there to be what are called critical periods of brain development ( <a class="user" href="http://en.wikipedia.org/wiki/Critical_period">http://en.wikipedia.org/wiki/Critical_period</a> ) when *specific areas* of the cortex *must* receive certain stimuli to develop properly, and if they don't, then you lose stereo vision (for example), and the rest of the cortex canNOT reprogram itself to restore it.