What is Digg?148 Comments
- Dumbledorito, on 10/12/2007, -1/+45"I used to think that the brain was the most fascinating part of the body. Then I realized, 'look who's telling me that.'"
- Emo Phillips - Dumbledorito, on 10/12/2007, -4/+48There are many people who make me think the average mechanized version of the human brain could be represented by a whack-a-mole game.
- godofpumpkins, on 10/12/2007, -7/+45well, unless we concede that there's some supernatural power guiding our brain, or an unpredictable (i.e., usefully random) subatomic behavior influencing our thought processes, then we should probably be able, with a sufficiently large computer, simulate the entire network of neurons, their excitement, chemical and electrical reactions between them, and so on. Not that such a massive parallel simulator is anywhere on the horizon, but basically:
If we can model the relevant structure of the brain, then we can simulate it. If we can simulate it, then the capabilities of the brain can be performed by a computer. If so, then the capabilities of a computer are a (possibly improper) superset of the capabilities of the brain. I'm talking about computer in the abstract sense (i.e., a UTM,) of course. - wonderchemist, on 10/12/2007, -1/+26Because we would like our pod bays doors to open when we what them to open.
- smartmlp, on 10/12/2007, -1/+18Because that would mean that they would have a penis.
- CGreen, on 10/12/2007, -2/+18@antiNeo
Neurons don't have "a few connections" they have thousands, neurons don't have "a bit of memory" and they don't make "decissions". They receive an input potential that creates an output potential there is no logic in them based on memory such as "this connection and this connection have this pattern, so now I send out that pattern on this output". The memory is built into the network.
They are in no way analogous to transistors in computers.
They are not just laid out in big mess, there is a clear structure with for example special built machinery for facial recognition (new born babies can recognize their mother after a few hours). - KlayBorg, on 10/12/2007, -10/+25We don't understand how the brain works well enough to be able to simulate it and make a computer act like it.
- godofpumpkins, on 10/12/2007, -2/+16Dammit my editing time ran out: Just to clarify, I'm not saying the brain is a model of a turing machine, or of any other known computational model, but simply that it can be simulated by one, and as such either solves the same set of problems computers can solve, or solves a subset of them.
- hartley, on 10/12/2007, -0/+12The question I keep asking myself is "Why can't my brain be more like a computer?"
- nepawoods, on 10/12/2007, -5/+17"the brain really IS a computer anyway"
Everything a computer can do can be completely explained without supposing the computer has any awareness of anything (thoughts, for instance) in the sense that we do. Therefore there is no reason to believe it has any such awareness, as we do. - rm999, on 10/12/2007, -1/+12I respect Jeff Hawkins and what he is doing, but in general the AI field does not care much for his work. His stuff is a simple addition to 20 year old techniques, and has not been shown to perform well except in toy examples.
His book "On Intelligence" was great from a philosophical perspective, but read it with a grain of salt. Thousands of researchers have learned that putting human-like intelligence into a computer is not a trivial task. - mercurysquad, on 10/12/2007, -2/+10Enough with the funny/witty comments!
Anyway, I believe a reasonable model of the human brain can be created using artificial neural networks. There are several different types of neural networks, training models, memory stores etc. which mimic specific patterns of real neural networks in the brain. Once we are actually able to map out the structure of the brain and the interconnections between them down to a sufficient level of detail, it should be possible to recreate an equivalent artificial network which approximates the behaviour of a real brain.
Of course it would need to be trained/taught and will learn and grow like a normal human first. Which means, we would need to provide about the same senses to it as us (although experimenting with different kinds of senses might be interesting). That means such an artificial brain might not be really as efficient or feasible (in the near future) as us humans, but the advantage might be that once a brain has been trained for a particular task, it can easily be cloned by transferring e.g. the connection weights or other parameters verbatim to a new model.
Of course in the end we might not even need a whole "brain".. just different smallers networks for particular tasks.
Just thoughts.. - GMorgan, on 10/12/2007, -0/+7Porn directly to your brain rather than via the inefficient optical interface.
- inactive, on 10/12/2007, -8/+14why cant it be more like a vagina? , the gift that keeps on giving
- PaulLev, on 10/12/2007, -2/+8There's no reason in principle why we couldn't build an artificial brain. We're just going about it the wrong way. Our brains evolved from the bottom up, from living organisms. Until we can make computers alive, we'll never be able to make them really think. http://www.paullevinson.net/archives/entries/books_by_paul.phtml#softedge
- sidd, on 10/12/2007, -0/+5oskite: The fact that DNA is quaternary has little bearing on the complexity of the things it can make. Anything in base4 is easily represented in base2. It is probably just lucky that DNA evolved as a quaternary system, rather than binary.
On the subject of the article: Few people seriously debate whether the brain could be simulated on a Turing machine, but it is a different story doing it in practice. The fact is that we are yet to discover many of the 'neat-tricks' of the brain. We haven't recreated human intelligence because we don't fully understand how it works. - beguiledfoil, on 10/12/2007, -1/+6Expanding on Anti's post:
Another issue when simulating organic brains is that the architectures are totally different. Brains are massively parallel architectures whereas the modern processor works primarily in serial. Even dual or quad core processors (or IBM's Cell) are essentially totally serial when compared with the embarrassingly parallel organic brain.
Calling Jeff Hawkin's idea a "software platform" seems a little disingenuous to me, the last time I read his group's stuff they were talking about little hierarchically-organized machines which would take in data and boil it down to conclusions. Seemed to me (an admitted laymen) to just add some specifics to the basic idea of a neural network. Obviously this could be simulated in software, but it would be ***** slow. - Erik1, on 10/12/2007, -2/+7@ adame
If you actually had a attention span longer than that of a goldfish, you would realize it is a good read. - CGreen, on 10/12/2007, -0/+4This research is not about creating a thinking machine. It's about creating a machine capable of learning. You do not think "that's a car" when you see one, but you have learned that it is a car and you remember it, this is what is being developed.
- MackDiesel2010, on 10/12/2007, -5/+9Hope: It is the quintessential human delusion, simultaneously the source of our greatest strength, and our greatest weakness.
- Robotsu, on 10/12/2007, -4/+8Why would it want to be? I thought the point of computers is that they did functions our brains are not well suited for??
- anothernobody, on 10/12/2007, -0/+4Exactly - I'm always wanting to google something in my head
- cactus476, on 10/12/2007, -3/+7Thinking like a human is not a standard. Thinking like a computer is.
They have their priorities wrong. - geronimo, on 10/12/2007, -2/+5There is only One being who could have done this and I'm sure you're thinking of who I am thinking of - that's right, The Flying Spaghetti Monster. Praise be upon Him.
- beguiledfoil, on 10/12/2007, -1/+4Yeah, no doubt! Humans have had electricity for a little over a hundred years now, clearly we should be emulating and improving on something that took hundreds of millions of years to evolve.
Your comment and your attitude are asinine, I hope you enjoy watching humanity leave you in its wake. - Sornos, on 10/12/2007, -1/+4A smart Windows would delete itself.
- GMorgan, on 10/12/2007, -1/+4Windows is bad because it is dumb, not because it is smart. A smart Windows would fix itself.
- Urusai, on 10/12/2007, -3/+6In case nobody has noticed, brains don't actually work very well. Making an artificial brain is like trying to grow an automobile from bacteria--you're gonna end up with a mess that doesn't really do what you want. You know, like humans.
- eddy144, on 10/12/2007, -5/+8Because our brain has been evolving for a damn long time. Computers have been around for 50 odd years. Problem solved.
- soapsuds, on 10/12/2007, -0/+3Einstein also wrote a glowing review of a book "disproving" the theory of plate tectonics. Newton spent the latter portion of his life pursuing alchemy. History is littered with people, brilliant in one field, overreaching their limitations in others.
Here is a refutation of Penrose's claim:
http://www.paul-almond.com/RefutationofPenroseGodelTuring.htm - Kyoushu, on 10/12/2007, -1/+4Oh, how terribly clever....
- geekee, on 10/12/2007, -1/+3Please open the pod bay doors HAL.
I can't do that, Dave. - Scynet, on 10/12/2007, -1/+3You either believe that consciousness/awareness is a result of chemical reactions and other natural events in our brain, or there's a deity involved. I really can't think of another way.
If they work through natural events, then we can eventually figure out our brains completely and are then able to recreate everything going on in our heads, including consciousness. Nature isn't "magic", deities are.
No? - KlayBorg, on 10/12/2007, -0/+2"In fact, there have been tremendous advances in brain research for years; but answers to the big questions are as elusive as ever."
http://www.spiegel.de/international/spiegel/0,1518,druck-466789,00.html - Ricapar, on 10/12/2007, -0/+2That's not how it happened. Get your story right.
Skynet first becomes self-aware and fires ze missiles. Many years pass, and -then- the search for Sarah Conner starts. - Dudeee, on 10/12/2007, -0/+2Reminds me of Terminator, were gonna be at war with machines.
- beguiledfoil, on 10/12/2007, -0/+2I'm afraid I couldn't relate the entire body of science and logic that encompasses my belief system in the bounds of this post, if you are actually interested in my beliefs e-mail me (see username)@gmail.com.
I can tell you what I don't believe, though: That an unknowably complicated and advance being (God) is a satisfactory explanation for the beginning of space and time, as well as the origin of all matter and energy in the universe. That is, to me, a non-answer. Rather than explain the beginning, the theory of a great creator merely moves the beginning back an unknowable amount.
And now I'm totally off-topic.. jerk. - GMorgan, on 10/12/2007, -1/+3I've yet to see a failsafe brain either.
- 4DFX, on 10/12/2007, -0/+2Why can't my brain be more like a computer... :P
- cwolves, on 10/12/2007, -0/+2There are already a group of scientists setting out to do just this, map a brain by measuring all the neurons. For the last 15 years they've been mapping the brain of a mouse and are almost done, hoping to switch the machine on in 2009-2010. There was a show on discovery or some similar channel about this recently.
IBM is doing the same thing: http://domino.watson.ibm.com/comm/pr.nsf/pages/rsc.bluegene_cognitive.html
There are already 100% working simulations of much simpler things like virii. - fedaykin3dfx, on 10/12/2007, -0/+2"Thou shalt not make a machine in the likeness of a human mind."
- godofpumpkins, on 10/12/2007, -0/+2you guys saying that we can't simulate it because there are millions of simultaneous operations going on in the brain are missing the point. Most arguments I hear against strong AI claim that there are fundamental differences, and that characteristics of the brain make it impossible to make something that behaves like it on a computer. My point is only that the brain is in fact computable, regardless of the immense machinery that we'd need to perform the task, and thus that there's nothing the brain can do that an abstract computer can't. To deny that is to say that we have a 'soul' (in any form) that _cannot_ be modeled, for if we could understand it and model it, we'd be able to simulate it.
- nepawoods, on 10/12/2007, -0/+2Here's another refutation: http://www.1729.com/consciousness/godel.html
Basically, Godel's Incompleteness theorem applies to consistent systems capable of representing an axiomatization of integer arithmetic. There is no evidence the brain is such a system. Furthermore, it concludes the system then is not complete. Well, there is no evidence the brain is complete (in that sense). - goettel, on 10/12/2007, -0/+1Easiest to quote (Wikipedia):
"HTM has been accused by the AI community of being nothing new, merely a rehash of many existing ideas, and without giving credit to their original authors. This is perhaps due to Hawkin's non-academic background and lack of peer reviews. The case is similar to that of Stephen Wolfram who similarly has made enough money in industry to bypass conventional academic publishing: reviews of both authors appeared back-to-back in the December 2005 issue of Artificial Intelligence." - TheKillDoctor, on 10/12/2007, -0/+1Here's a possibility:
It seems it is more likely that the merging of technology with our own brains and senses will just make us this dreamed about artificial intelligent life form. Yeah I'm being a little sarcastic but seriously the merging of man and machine has a higher probability of eventually spawning AI in oh lets say the next 100 years. Just a another random thought to throw out here...... - PsychoticClown, on 10/12/2007, -2/+3G Morgan:
>>"If I was designed intelligently then why do I not have an ALU built into my brain. Sounds more like stupid design to me." - mrfunkeye, on 10/12/2007, -1/+2IMHO the human brain could never be simulated by a computer since one could never tell which is the ueber-parent-process inside our brain and how it is forking its child-processes.
Furthermore, with the knowledge we have today we cannot even build a failsafe OS, so simulating the human brain on a non-perfect computer is a no-brainer. - GMorgan, on 10/12/2007, -1/+2If I was designed intelligently then why do I not have an ALU built into my brain. Sounds more like stupid design to me.
- mark101, on 10/12/2007, -0/+1A brain has motivation built into it.To get away from pain and feel pleasure, feel hunger >get food, feel tired > sleep , feel angry> fight, etc. What does the computer feel to motivate it?
- BogogF, on 10/12/2007, -0/+1"My point is that creation requires a creator. We are here, so there must be a creator. It doesn't matter what we know about it/him. If we see graffiti on the wall we know someone did it. We know nothing about this person or where he came form but we know graffiti doesn't just create itself. Same for the universe, which is - needless to say - infinitely more complex than graffiti."
Creation requires a creator, explain how the creator was created. It does matter, we must know to form a conclusion. Graffiti comes from a human manufacturing process, ie we create it... we know this for a fact, god we don't know for a fact, however evolution is more believable. Intelligent design is just the rewording of the same ***** that's been said time and time again, our world is here, god must have created it. Which isn't really saying anything intelligent at all, it's just a regurgitation of what you lot have always said but it's worded in such a way as to confuse, and we all know what confusion causes, uncertainty and doubt.
I'm quite happy to live my life knowing that this is the only shot I get and when I die I stop existing, my body turns to useless meat and I don't think anymore... it makes me appreciate life and my surroundings. I may never get to see this breed of moth, this rain drop, these clouds ever again, forever. Whereas three quarters of the world believes they'll live forever... no wonder the world is being squandered and tens of thousands of people die everyday NEEDLESSLY for some backwater ***** patch of ***** dirt that in the end doesn't matter, because you don't exist anymore. Not to mention the ***** greedy pigs that run the world. -
Show 51 - 100 of 147 discussions
The Digg Toolbar for Firefox lets you Digg, submit content, and keep track of Digg even when you're not on the Digg site. Download the official 
© Digg Inc. 2009
— Content created and posted by Digg users is