RE: [agi] IBM, Los Alamos scientists claim fastest computer
Matt, Thank you for your reply. For me it is very thought provoking. -Original Message- From: Matt Mahoney [mailto:[EMAIL PROTECTED] Sent: Thursday, June 12, 2008 7:23 PM To: agi@v2.listbox.com Subject: RE: [agi] IBM, Los Alamos scientists claim fastest computer --- On Thu, 6/12/08, Ed Porter [EMAIL PROTECTED] wrote: I think processor to memory, and inter processor communications are currently far short Your concern is over the added cost of implementing a sparsely connected network, which slows memory access and requires more memory for representation (e.g. pointers in addition to a weight matrix). We can alleviate much of the problem by using connection locality. [Ed Porter] this would certainly be true if it worked The brain has about 10^11 neurons with 10^4 synapses per neuron. If we divide this work among 10^6 processors, each representing 1 mm^3 of brain tissue, then each processor must implement 10^5 neurons and 10^9 synapses. By my earlier argument, there can be at most 10^6 external connection assuming 1-2 micron nerve fiber diameter, [Ed Porter] -- Why couldn't each of the 10^6 fibers have multiple connections along its length within the cm^3 (although it could be represented as one row in the matrix, with individual connections represented as elements in such a row) so half of the connections must be local. This is true at any scale because when you double the size of a cube, you increase the number of neurons by 8 but increase the number of external connections by 4. Thus, for any size cube, half of the external connections are to neighboring cubes and half are to more distant cubes. [Ed Porter] -- I am getting lost here. Why are half the connections local. You implied there are 10^6 external connections in the cm^3, and 10^9 synapses, which are the connections. Thus the 10^6 external connections you mention are only 1/1000 of 10^9 total connections you mention in the cm^3, not one half as you say. I understand that there are likely to be as many connections leaving the cube as going into it, which is related, but not that same thing as saying half the connection in the cm^3 are external. [Ed Porter] -- It is true that the rate of change for each doubling in scale of the ratio surface to volume remains 1/2, but the actual ratio of surface to volume decreases by 1/2 at each such doubling of scale, meaning the ratio actually DOES CHANGE with scaling, rather remaining constant, as indicated above. A 1 mm^3 cube can be implemented as a fully connected 10^5 by 10^5 matrix of 10^10 connections. This could be implemented as a 1.25 GB array of bits with 5% of bits set to 1 representing a connection. [Ed Porter] a synaps would have multiple weights, such as short term and long term weights and they would each be more than one bit. Plus some are excitatory and others or inhibitory, so they would have differing signs. So multiple bits, probably at least two bytes would be necessary per element in the matrix. [Ed Porter] -- Also you haven't explained how you efficiently do the activation between cubes (I assume it would be by having a row for each neuron that projects axons into the cube, and a column for each neuron that projects a dendrite into it). This could still be represented by the matrix, but it would tend to increase its sparseness. [Ed Porter] -- Learning changes in which dendrites and axons projected into a cube would require changing the matrix, which is doable, but can make things more complicated. Another issue is how many other cubes would each cm^3 communicate with. Are we talking 10, 100, 10^3, 10^4, 10^5, or 10^6. The number could have a significant impact on communication costs. [Ed Porter] -- I don't think this system would be good for my current model for AGI representation, which is based on a type of graph matching, rather than just a simple summing of synaptic inputs. The internal computation bottleneck is the vector product which would be implemented using 128 bit AND instructions in SSE2 at full serial memory bandwidth. External communication is at most one bit per connected neuron every cycle (20-100 ms), because the connectivity graph does not change rapidly. A randomly connected sparse network could be described compactly using hash functions. [Ed Porter] -- It is interesting to think that this actually could be used to speed up the processing of simple neural models. I understand how the row values associated with the axon synapses of a given neuron could be read rapidly in a serial manner. And how run-length encoding, or some other means could be used to more compactly represent a sparse matrix. I also understand how the contributions for to the activation for each of the 10^5 columns made by each row could be stored in L2 cache at a rate of about 100mHz. [Ed Porter] -- L2 Cache write commonly take about 10 to 20clock cycles. --- perhaps you could write them into memory blocks in L1 cache, which might only take about two
Re: [agi] Nirvana
There've been enough responses to this that I will reply in generalities, and hope I cover everything important... When I described Nirvana attractractors as a problem for AGI, I meant that in the sense that they form a substantial challenge for the designer (as do many other features/capabilities of AGI!), not that it was an insoluble problem. The hierarchical fixed utility function is probably pretty good -- not only does it match humans (a la Maslow) but Asimov's Three Laws. And it can be more subtle than it originally appears: Consider a 3-Laws robot that refuses to cut a human with a knife because that would harm her. It would be unable to become a surgeon, for example. But the First Law has a clause, or through inaction allow a human to come to harm, which means that the robot cannot obey by doing nothing -- it must weigh the consequences of all its possible courses of action. Now note that it hasn't changed its utility function -- it always believed that, say, appendicitis is worse than an incision -- but what can happen is that its world model gets better and it *looks like* it's changed its utility function because it now knows that operations can cure appendicitis. Now it seems reasonable that this is a lot of what happens with people, too. And you can get a lot of mileage out of expressing the utility function in very abstract terms, e.g. life-threatening disease so that no utility function update is necessary when you learn about a new disease. The problem is that the more abstract you make the concepts, the more the process of learning an ontology looks like ... revising your utility function! Enlightenment, after all, is a Good Thing, so anything that leads to it, nirvana for example, must be good as well. So I'm going to broaden my thesis and say that the nirvana attractors lie in the path of *any* AI with unbounded learning ability that creates new abstractions on top of the things it already knows. How to avoid them? I think one very useful technique is to start with the kind of knowledge and introspection capability to let the AI know when it faces one, and recognize that any apparent utility therein is fallacious. Of course, none of this matters till we have systems that are capable of unbounded self-improvement and abstraction-forming, anyway. Josh --- agi Archives: http://www.listbox.com/member/archive/303/=now RSS Feed: http://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: http://www.listbox.com/member/?member_id=8660244id_secret=103754539-40ed26 Powered by Listbox: http://www.listbox.com
Re: [agi] Nirvana
Most people are about as happy as they make up their minds to be. -- Abraham Lincoln In our society, after a certain point where we've taken care of our immediate needs, arguably we humans are and should be subject to the Nirvana effect. Deciding that you can settle for something (if your subconscious truly can handle it) definitely makes you more happy than not. If, like a machine, you had complete control over your subconscious/utility functions, you *could* Nirvana yourself by happily accepting anything. This is why pleasure and lack of pain suck as goals. They are not goals, they are status indicators. If you accept them as goals, nirvana is clearly the fastest, cleanest, and most effective way to fulfill them. Why is this surprising or anything to debate about? - Original Message - From: J Storrs Hall, PhD [EMAIL PROTECTED] To: agi@v2.listbox.com Sent: Friday, June 13, 2008 11:58 AM Subject: Re: [agi] Nirvana There've been enough responses to this that I will reply in generalities, and hope I cover everything important... When I described Nirvana attractractors as a problem for AGI, I meant that in the sense that they form a substantial challenge for the designer (as do many other features/capabilities of AGI!), not that it was an insoluble problem. The hierarchical fixed utility function is probably pretty good -- not only does it match humans (a la Maslow) but Asimov's Three Laws. And it can be more subtle than it originally appears: Consider a 3-Laws robot that refuses to cut a human with a knife because that would harm her. It would be unable to become a surgeon, for example. But the First Law has a clause, or through inaction allow a human to come to harm, which means that the robot cannot obey by doing nothing -- it must weigh the consequences of all its possible courses of action. Now note that it hasn't changed its utility function -- it always believed that, say, appendicitis is worse than an incision -- but what can happen is that its world model gets better and it *looks like* it's changed its utility function because it now knows that operations can cure appendicitis. Now it seems reasonable that this is a lot of what happens with people, too. And you can get a lot of mileage out of expressing the utility function in very abstract terms, e.g. life-threatening disease so that no utility function update is necessary when you learn about a new disease. The problem is that the more abstract you make the concepts, the more the process of learning an ontology looks like ... revising your utility function! Enlightenment, after all, is a Good Thing, so anything that leads to it, nirvana for example, must be good as well. So I'm going to broaden my thesis and say that the nirvana attractors lie in the path of *any* AI with unbounded learning ability that creates new abstractions on top of the things it already knows. How to avoid them? I think one very useful technique is to start with the kind of knowledge and introspection capability to let the AI know when it faces one, and recognize that any apparent utility therein is fallacious. Of course, none of this matters till we have systems that are capable of unbounded self-improvement and abstraction-forming, anyway. Josh --- agi Archives: http://www.listbox.com/member/archive/303/=now RSS Feed: http://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: http://www.listbox.com/member/?; Powered by Listbox: http://www.listbox.com --- agi Archives: http://www.listbox.com/member/archive/303/=now RSS Feed: http://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: http://www.listbox.com/member/?member_id=8660244id_secret=103754539-40ed26 Powered by Listbox: http://www.listbox.com
Re: [agi] Nirvana
In my visualization of the Cosmic All, it is not surprising. However, there is an undercurrent of the Singularity/AGI community that is somewhat apocaliptic in tone, and which (to my mind) seems to imply or assume that somebody will discover a Good Trick for self-improving AIs and the jig will be up with the very first one. I happen to think it'll be a lot more like the Industrial Revolution -- it'll take a lot of work by a lot of people, but revolutionary in its implications for the human condition even so. I'm just trying to point out where I think some of the work will have to go. I think that our culture of self-indulgence is to some extent in a Nirvana attractor. If you think that's a good thing, why shouldn't we all lie around with wires in our pleasure centers (or hopped up on cocaine, same difference) with nutrient drips? I'm working on AGI because I want to build a machine that can solve problems I can't do alone. The really important problems are not driving cars, or managing companies, or even curing cancer, although building machines that can do these things will be of great benefit. The hard problems are moral ones, how to live in increasingly complex societies without killing each other, and so forth. That's why it matters that an AGI be morally self-improving as well as intellectually. pax vobiscum, Josh On Friday 13 June 2008 12:29:33 pm, Mark Waser wrote: Most people are about as happy as they make up their minds to be. -- Abraham Lincoln In our society, after a certain point where we've taken care of our immediate needs, arguably we humans are and should be subject to the Nirvana effect. Deciding that you can settle for something (if your subconscious truly can handle it) definitely makes you more happy than not. If, like a machine, you had complete control over your subconscious/utility functions, you *could* Nirvana yourself by happily accepting anything. This is why pleasure and lack of pain suck as goals. They are not goals, they are status indicators. If you accept them as goals, nirvana is clearly the fastest, cleanest, and most effective way to fulfill them. Why is this surprising or anything to debate about? --- agi Archives: http://www.listbox.com/member/archive/303/=now RSS Feed: http://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: http://www.listbox.com/member/?member_id=8660244id_secret=106510220-47b225 Powered by Listbox: http://www.listbox.com
RE: [agi] IBM, Los Alamos scientists claim fastest computer
--- On Fri, 6/13/08, Ed Porter [EMAIL PROTECTED] wrote: [Ed Porter] -- Why couldn't each of the 10^6 fibers have multiple connections along its length within the cm^3 (although it could be represented as one row in the matrix, with individual connections represented as elements in such a row) I think you mean 10^6 fibers in 1 cubic millimeter. They would have multiple connections, but I am only counting interprocessor communication, which is 1 bit to transmit the state of the neuron (on or off) or a few bits to transmit its activation level to neighboring processors. With regard to representing different types of synapses (various time delays, strength bounds, learning rates, etc), this information can be recorded as characteristics of the input and output neurons and derived as needed to save space. Minimizing inter-processor communication is a harder problem. This can be done by mapping the neural network into a hierarchical organization so that groups of co-located neurons are forced to communicate with other groups through narrow channels using a small number of neurons. We know that many problems can be solve this way. For example, a semantic language model made of a 20K by 20K word association matrix can be represented using singular value decomposition as a 3 layer neural network with about 100 to 200 hidden neurons [1,2]. The two weight matrices could then be implemented on separate processors which communicate through the hidden layer neurons. More generally, we know from chaos theory that complex systems must limit the number of interconnections to be stable [3], which suggests that many AI problems in general can be decomposed this way. Remember we need not model the human brain in precise detail, since our goal is to solve AGI by any means. We are allowed to use more efficient algorithms if we discover them. I ran some benchmarks on my PC (2.2 GHz Athlon-64, 3500+). It copies large arrays at 1 GB per second using MMX or SSE2, which is not quite fast enough for a 10^5 by 10^5 neural network simulation. 1. Bellegarda, Jerome R., John W. Butzberger, Yen-Lu Chow, Noah B. Coccaro, Devang Naik (1996), “A novel word clustering algorithm based on latent semantic analysis”, Proc. IEEE Intl. Conf. on Acoustics, Speech, and Signal Processing, vol. 1, 172-175. 2. Gorrell, Genevieve (2006), “Generalized Hebbian Algorithm for Incremental Singular Value Decomposition in Natural Language Processing”, Proceedings of EACL 2006, Trento, Italy. http://www.aclweb.org/anthology-new/E/E06/E06-1013.pdf 3. Kauffman, Stuart A. (1991), “Antichaos and Adaptation”, Scientific American, Aug. 1991, p. 64. -- Matt Mahoney, [EMAIL PROTECTED] --- agi Archives: http://www.listbox.com/member/archive/303/=now RSS Feed: http://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: http://www.listbox.com/member/?member_id=8660244id_secret=106510220-47b225 Powered by Listbox: http://www.listbox.com
Re: [agi] Nirvana
I think that our culture of self-indulgence is to some extent in a Nirvana attractor. If you think that's a good thing, why shouldn't we No, I think it's a bad thing. That's why I said This is why pleasure and lack of pain suck as goals. However, there is an undercurrent of the Singularity/AGI community that is somewhat apocaliptic in tone, Yeah, well, I would (and will, shortly) argue differently. - Original Message - From: J Storrs Hall, PhD [EMAIL PROTECTED] To: agi@v2.listbox.com Sent: Friday, June 13, 2008 1:28 PM Subject: Re: [agi] Nirvana In my visualization of the Cosmic All, it is not surprising. However, there is an undercurrent of the Singularity/AGI community that is somewhat apocaliptic in tone, and which (to my mind) seems to imply or assume that somebody will discover a Good Trick for self-improving AIs and the jig will be up with the very first one. I happen to think it'll be a lot more like the Industrial Revolution -- it'll take a lot of work by a lot of people, but revolutionary in its implications for the human condition even so. I'm just trying to point out where I think some of the work will have to go. I think that our culture of self-indulgence is to some extent in a Nirvana attractor. If you think that's a good thing, why shouldn't we all lie around with wires in our pleasure centers (or hopped up on cocaine, same difference) with nutrient drips? I'm working on AGI because I want to build a machine that can solve problems I can't do alone. The really important problems are not driving cars, or managing companies, or even curing cancer, although building machines that can do these things will be of great benefit. The hard problems are moral ones, how to live in increasingly complex societies without killing each other, and so forth. That's why it matters that an AGI be morally self-improving as well as intellectually. pax vobiscum, Josh On Friday 13 June 2008 12:29:33 pm, Mark Waser wrote: Most people are about as happy as they make up their minds to be. -- Abraham Lincoln In our society, after a certain point where we've taken care of our immediate needs, arguably we humans are and should be subject to the Nirvana effect. Deciding that you can settle for something (if your subconscious truly can handle it) definitely makes you more happy than not. If, like a machine, you had complete control over your subconscious/utility functions, you *could* Nirvana yourself by happily accepting anything. This is why pleasure and lack of pain suck as goals. They are not goals, they are status indicators. If you accept them as goals, nirvana is clearly the fastest, cleanest, and most effective way to fulfill them. Why is this surprising or anything to debate about? --- agi Archives: http://www.listbox.com/member/archive/303/=now RSS Feed: http://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: http://www.listbox.com/member/?; Powered by Listbox: http://www.listbox.com --- agi Archives: http://www.listbox.com/member/archive/303/=now RSS Feed: http://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: http://www.listbox.com/member/?member_id=8660244id_secret=106510220-47b225 Powered by Listbox: http://www.listbox.com
Re: [agi] Nirvana
On Fri, Jun 13, 2008 at 1:28 PM, J Storrs Hall, PhD [EMAIL PROTECTED] wrote: I think that our culture of self-indulgence is to some extent in a Nirvana attractor. If you think that's a good thing, why shouldn't we all lie around with wires in our pleasure centers (or hopped up on cocaine, same difference) with nutrient drips? Because it's unsafe for now. We will eventually work it out. Jiri --- agi Archives: http://www.listbox.com/member/archive/303/=now RSS Feed: http://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: http://www.listbox.com/member/?member_id=8660244id_secret=106510220-47b225 Powered by Listbox: http://www.listbox.com
Re: [agi] Nirvana
Mark, Assuming that a) pain avoidance and pleasure seeking are our primary driving forces; and b) our intelligence wins over our stupidity; and c) we don't get killed by something we cannot control; Nirvana is where we go. Jiri --- agi Archives: http://www.listbox.com/member/archive/303/=now RSS Feed: http://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: http://www.listbox.com/member/?member_id=8660244id_secret=106510220-47b225 Powered by Listbox: http://www.listbox.com
Re: [agi] Nirvana
Yes, but I strongly disagree with assumption one. Pain avoidance and pleasure are best viewed as status indicators, not goals. - Original Message - From: Jiri Jelinek [EMAIL PROTECTED] To: agi@v2.listbox.com Sent: Friday, June 13, 2008 3:42 PM Subject: Re: [agi] Nirvana Mark, Assuming that a) pain avoidance and pleasure seeking are our primary driving forces; and b) our intelligence wins over our stupidity; and c) we don't get killed by something we cannot control; Nirvana is where we go. Jiri --- agi Archives: http://www.listbox.com/member/archive/303/=now RSS Feed: http://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: http://www.listbox.com/member/?; Powered by Listbox: http://www.listbox.com --- agi Archives: http://www.listbox.com/member/archive/303/=now RSS Feed: http://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: http://www.listbox.com/member/?member_id=8660244id_secret=106510220-47b225 Powered by Listbox: http://www.listbox.com
Re: [agi] The Logic of Nirvana
a future AGI will probably believe one thing, but act as if it believes something quite different, for very logical reasons. I wish/hope we can avoid that. AGI should IMO follow scientific principles. If its honesty hurts then our society/environment should be targeted for a change. Regards, Jiri Jelinek --- agi Archives: http://www.listbox.com/member/archive/303/=now RSS Feed: http://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: http://www.listbox.com/member/?member_id=8660244id_secret=106510220-47b225 Powered by Listbox: http://www.listbox.com
Re: [agi] The Logic of Nirvana
On Friday 13 June 2008 02:42:10 pm, Steve Richfield wrote: Buddhism teaches that happiness comes from within, so stop twisting the world around to make yourself happy, because this can't succeed. However, it also teaches that all life is sacred, so pay attention to staying healthy. In short, attend to the real necessities and don't sweat the other stuff. A better example of goal abstraction I couldn't have made up myself. --- agi Archives: http://www.listbox.com/member/archive/303/=now RSS Feed: http://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: http://www.listbox.com/member/?member_id=8660244id_secret=106510220-47b225 Powered by Listbox: http://www.listbox.com
Re: [agi] Nirvana
a) pain avoidance and pleasure seeking are our primary driving forces; On Fri, Jun 13, 2008 at 3:47 PM, Mark Waser [EMAIL PROTECTED] wrote: Yes, but I strongly disagree with assumption one. Pain avoidance and pleasure are best viewed as status indicators, not goals. Pain and pleasure [levels] might be indicators (or primary action triggers), but I think it's ok to call pain avoidance and pleasure seeking our driving forces. I cannot think of any intentional human activity which is not somehow associated with those primary triggers/driving forces and that's why I believe the assumption one is valid. Best, Jiri --- agi Archives: http://www.listbox.com/member/archive/303/=now RSS Feed: http://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: http://www.listbox.com/member/?member_id=8660244id_secret=106510220-47b225 Powered by Listbox: http://www.listbox.com
Re: [agi] The Logic of Nirvana
Buddhism teaches that happiness comes from within, so stop twisting the world around to make yourself happy, because this can't succeed. Which is of course false... It might come within but triggers can be internal as well as external and both work pretty well. For the world twisting, it's just a matter of having enough power = works well for fewer. Jiri Religion is all bunk. [Thomas A. Edison] --- agi Archives: http://www.listbox.com/member/archive/303/=now RSS Feed: http://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: http://www.listbox.com/member/?member_id=8660244id_secret=106510220-47b225 Powered by Listbox: http://www.listbox.com
Re: [agi] Nirvana
Your belief value is irrelevant to reality. Of course all human activity is associated with pain and pleasure because evolution gave us pleasure and pain to motivate us to do smart things (as far as evolution is concerned) and avoid stupid things (and yes, I am anthropomorphizing evolution for ease of communication but if you can't figure out what I really mean . . . . ). However, correlation is not equivalent to causation. Goal is survival or propagation of species. Evolution rewards or punishes according to these goals. If you ignore these goals and reprogram your pleasure and pain you go extinct. More clearly, if you wire-head, you go extinct (i.e. you are an evolutionary loser). Go ahead and wirehead if you wish but don't be surprised if someone with the same values decides that he is allowed to kill you painlessly since you're eating up their resources to promote their own pleasure. But then again, it really doesn't matter because you're extinct either way, right? - Original Message - From: Jiri Jelinek [EMAIL PROTECTED] To: agi@v2.listbox.com Sent: Friday, June 13, 2008 4:34 PM Subject: Re: [agi] Nirvana a) pain avoidance and pleasure seeking are our primary driving forces; On Fri, Jun 13, 2008 at 3:47 PM, Mark Waser [EMAIL PROTECTED] wrote: Yes, but I strongly disagree with assumption one. Pain avoidance and pleasure are best viewed as status indicators, not goals. Pain and pleasure [levels] might be indicators (or primary action triggers), but I think it's ok to call pain avoidance and pleasure seeking our driving forces. I cannot think of any intentional human activity which is not somehow associated with those primary triggers/driving forces and that's why I believe the assumption one is valid. Best, Jiri --- agi Archives: http://www.listbox.com/member/archive/303/=now RSS Feed: http://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: http://www.listbox.com/member/?; Powered by Listbox: http://www.listbox.com --- agi Archives: http://www.listbox.com/member/archive/303/=now RSS Feed: http://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: http://www.listbox.com/member/?member_id=8660244id_secret=106510220-47b225 Powered by Listbox: http://www.listbox.com
Re: [agi] Nirvana
On Fri, Jun 13, 2008 at 6:21 PM, Mark Waser [EMAIL PROTECTED] wrote: if you wire-head, you go extinct Doing it today certainly wouldn't be a good idea, but whatever we do to take care of risks and improvements, our AGI(s) will eventually do a better job, so why not then? Regards, Jiri Jelinek --- agi Archives: http://www.listbox.com/member/archive/303/=now RSS Feed: http://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: http://www.listbox.com/member/?member_id=8660244id_secret=106510220-47b225 Powered by Listbox: http://www.listbox.com