Re: [agi] Pure reason is a disease.
Hi Mark, AGI(s) suggest solutions people decide what to do. 1. People are stupid and will often decide to do things that will kill large numbers of people. I wonder how vague are the rules used by major publishers to decide what is OK to publish. I'm proposing a layered defense strategy Force the malevolent individual to navigate multiple defensive layers and you better the chances of detecting and stopping him. Can you get more specific about the layers? How do you detect malevolent individuals? Note that the fact that a particular user is highly interested in malevolent stuff doesn't mean he is bad guy. 2. The AGI will, regardless of what you do, fairly shortly be able to take actions on it's own. Without feelings, it cannot prefer = won't do a thing on its own. More powerful problem solver - Sure. The ultimate decision maker - I would not vote for that. The point is -- you're not going to get a vote. It's going to happen whether you like it or not. Unless we mess up, our machines do what we want. I don't think we necessarily have to mess up. The fact that the AGI can keep digging through (and keep fixing) its data very systematically doesn't solve the time constraint and deadline problems. Sure, there will be limitations. But if an AGI gets a) start scenario b) target scenario c) User-provided rules to follow d) System-config based rules to follow (e.g don't use knowledge marked [security_marking] when generation solutions for members of 'user_role_name' role) e) deadline then it can just show the first valid solution found, or say something like Sorry, can't make it + a reason (e.g. insufficient knowledge/time or thought broken by info access restriction) And multiple layers of defense make it harder to hack. Your arguments conflict with each other. When talking about hacking, I meant unauthorized access and/or modifications of AGI's resources. Considering current technology, there are many standard ways for multi-layer security. When it comes to generating safe system responses to regular user-requests then see above. Being busy with the knowledge representation issues, I did not figure out the exact implementation of the security marking algorithm yet. It might get tricky and I don't think I'll find practical hints in emotions. To some extent it might be handled by selected users. Look at it this way. Your logic says that if you can build this perfect shining AGI on a hill -- that everything will be OK. My emotions say that there is far too much that can go awry if you depend upon *everything* that you say you're depending upon *plus* everything that you don't realize you're depending upon *plus* . . . Playing with powerful tools always includes risks. More and more powerful tools will be developed. If we cannot deal with it then we don't deserve future. But I'm optimistic. Hopefully, AGI will get it right when asked to help us to figure out how to make sure we deserve it ;-) Regards, Jiri Jelinek On 5/16/07, Mark Waser [EMAIL PROTECTED] wrote: AGI should IMO focus on a) figuring out how to reach given goals, instead of b) trying to guess if users want something else than what they actually asked for. Absotively, positutely. b) is a recipe for disaster and my biggest gripe with Eliezer Yudkowski and the SIAI. What is unsafe to show sometimes depends on the level of details Figuring out the safe level of detail is not always easy and another problem is that smart users could break malevolent goals into separate tasks so that [at least the first generation] AGIs wouldn't be able to detect it even when following your emotion-related rules. The users could be using multiple accounts so even if all those tasks are given to a single instance of an AGI, it might not be able to notice the master plan. So is it dangerous? Sure, it is.. Yes, nothing is fool-proof given a sufficiently talented fool. That's why I'm proposing a layered defense strategy. Don't allow a single point of failure before the world goes kaboom! Force the malevolent individual to navigate multiple defensive layers and you better the chances of detecting and stopping him. AGI is potentially very powerful tool, but what we do with it is up to us. Nope. It's up to the eight billion plus morons who will access it. Actually, it's up to itself when some really bright fool modifies it in a certain way. 2. non-optimally stored and integrated knowledge Then you want to fix the cause by optimizing integrating instead of solving symptoms by adding backup searches. You clearly don't get the operating in the limited dirty time-constrained world thing. Building knowledge in the real world always leaves a trail of incomplete and unintegrated knowledge. Yes, the builder always follows behind and gathers more knowledge and integrates better -- but the real world also includes time constraints and deadlines for action. This isn't AIXI we're talking about. In a perfect world,
[agi] Write a doctoral dissertation, trigger a Singularity
University graduate students in computer science, linguistics, psychology, neuroscience and so on need a suitable topic for that scholarly contribution known as a Ph.D. dissertation. The SourceForge Mind project in artificial intelligence, on the other hand, needs entree into the academic AI literature. Why not start your academic career with a blockbuster dissertation? Think back to Erwin Schrodinger writing his equation in 1926. He got the idea from a French physicist who had recently defended his doctoral dissertation. Talk about the impact of a dissertation -- it literally had the bang of an atomic bomb. Your impact, Dr. Science, could be even greater. Your Ph.D. thesis could trigger the doomsday scenario of the Technological Singularity. The Singularity is near, but it hasn't happened yet because you have not yet submitted your bestseller-book-quality dissertation. CRITICAL MASS The AI Manhattan Project will not get off the ground until we assemble a scientific infrastructure of experts trained in the theory and practice of constructing artificially intelligent minds for robots. A few prototypes such as http://AIMind-I.com and http://mind.sourceforge.net/Mind.html are already out there, but we need a pre-Cambrian explosion of virally proliferating Mind versions if there is to be a Darwinian eco-system of AI Minds racing through the Internet and engaging in the mortal competition ending in the survival of the fittest. Your book qua Ph.D. dissertation may suceed where AI4U has failed -- as the Gutenberg Bible of the arrival of True AI-Complete. You may start by simply publishing a few scholarly papers on open-source artificial intelligence. Magazine and newspaper articles may flow from you, but the real target is academia. You are permitted -- and in fact it is your duty -- to take a critical stance towards the extraordinary scientific claims made when the Mind project asserts that AI has been solved, but you should shy away from embarassing yourself through woefully ignorant Mentifex-bashing such as happened with the http://www.advogato.org/article/928.html Advogato Has Failed debacle, where the author could not himself discredit Mentifex and so he ignorantly cited two attacks on Mentifex that were actually written by one and the same Internet cyberstalker. We want here a growing tree of scientific illumination, not a chain of thoughtless me-too ad-hominem sniper attacks. Above all avoid the endless, non-productive jawboning about artificial intelligence such as occurs year in and year out at http://www.mail-archive.com/agi@v2.listbox.com and other forums where blowhard discussants quibble about the AI climate but never write any code that advances the state of the art. So develop a thesis and run it by your faculty advisor. Stay aloof from the Mind project to keep your independence. When the facts are in and your case is made, publish and become a Philosophiae Doctor -- a teacher of philosophy. - This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?member_id=231415user_secret=fabd7936
Re: [agi] Pure reason is a disease.
I wonder how vague are the rules used by major publishers to decide what is OK to publish. Generally, there are no rules -- it's normally just the best judgment of a single individual. Can you get more specific about the layers? How do you detect malevolent individuals? Note that the fact that a particular user is highly interested in malevolent stuff doesn't mean he is bad guy. Sure. There's the logic layer and the emotion layer. Even if the logic layer get convinced, the emotion layer is still there to say Whoa. Hold on a minute. Maybe I'd better run this past some other people . . . . Note also, I'm not trying to detect a malevolent individual. I'm trying to prevent facilitating an action that could be harmful. I don't care about whether the individual is malevolent or stupid (though, in later stages, malevolence detection probably would be a good idea so as to possibly deny the user unsupervised access to the system). Without feelings, it cannot prefer = won't do a thing on its own. Nope. Any powerful enough system is going to have programmed goals which it then will have to interpret and develop subgoals and a plan of action. While it may not have set the top-level goal(s), it certainly is operating on it's own. Unless we mess up, our machines do what we want. I don't think we necessarily have to mess up. We don't have to necessarily mess up. I can walk a high-wire if you give me two hand-rails. But not putting the hand-rails in place would be suicide for me. c) User-provided rules to follow The crux of the matter. Can you specify rules that won't conflict with each other and which cover every contingency? If so, what is the difference between them and an unshakeable attraction or revulsion? Mark - Original Message - From: Jiri Jelinek [EMAIL PROTECTED] To: agi@v2.listbox.com Sent: Sunday, May 20, 2007 4:14 AM Subject: Re: [agi] Pure reason is a disease. Hi Mark, AGI(s) suggest solutions people decide what to do. 1. People are stupid and will often decide to do things that will kill large numbers of people. I wonder how vague are the rules used by major publishers to decide what is OK to publish. I'm proposing a layered defense strategy Force the malevolent individual to navigate multiple defensive layers and you better the chances of detecting and stopping him. Can you get more specific about the layers? How do you detect malevolent individuals? Note that the fact that a particular user is highly interested in malevolent stuff doesn't mean he is bad guy. 2. The AGI will, regardless of what you do, fairly shortly be able to take actions on it's own. Without feelings, it cannot prefer = won't do a thing on its own. More powerful problem solver - Sure. The ultimate decision maker - I would not vote for that. The point is -- you're not going to get a vote. It's going to happen whether you like it or not. Unless we mess up, our machines do what we want. I don't think we necessarily have to mess up. The fact that the AGI can keep digging through (and keep fixing) its data very systematically doesn't solve the time constraint and deadline problems. Sure, there will be limitations. But if an AGI gets a) start scenario b) target scenario c) User-provided rules to follow d) System-config based rules to follow (e.g don't use knowledge marked [security_marking] when generation solutions for members of 'user_role_name' role) e) deadline then it can just show the first valid solution found, or say something like Sorry, can't make it + a reason (e.g. insufficient knowledge/time or thought broken by info access restriction) And multiple layers of defense make it harder to hack. Your arguments conflict with each other. When talking about hacking, I meant unauthorized access and/or modifications of AGI's resources. Considering current technology, there are many standard ways for multi-layer security. When it comes to generating safe system responses to regular user-requests then see above. Being busy with the knowledge representation issues, I did not figure out the exact implementation of the security marking algorithm yet. It might get tricky and I don't think I'll find practical hints in emotions. To some extent it might be handled by selected users. Look at it this way. Your logic says that if you can build this perfect shining AGI on a hill -- that everything will be OK. My emotions say that there is far too much that can go awry if you depend upon *everything* that you say you're depending upon *plus* everything that you don't realize you're depending upon *plus* . . . Playing with powerful tools always includes risks. More and more powerful tools will be developed. If we cannot deal with it then we don't deserve future. But I'm optimistic. Hopefully, AGI will get it right when asked to help us to figure out how to make sure we deserve it ;-) Regards, Jiri Jelinek On 5/16/07, Mark Waser [EMAIL
[agi] Relationship btw consciousness and intelligence
Hi all, Someone emailed me recently about Searle's Chinese Room argument, http://en.wikipedia.org/wiki/Chinese_room a topic that normally bores me to tears, but it occurred to me that part of my reply might be of interest to some on this list, because it pertains to the more general issue of the relationship btw consciousness and intelligence. It also ties in with the importance of thinking about efficient intelligence rather than just raw intelligence, as discussed in the recent thread on definitions of intelligence. Here is the relevant part of my reply about Searle: However, a key point is: The scenario Searle describes is likely not physically possible, due to the unrealistically large size of the rulebook. The structures that we associate with intelligence (will, focused awareness, etc.) in a human context, all come out of the need to do intelligent processing within modest space and time requirements. So when we say we feel like the {Searle+rulebook} system isn't really understanding Chinese, what we mean is: It isn't understanding Chinese according to the methods we are used to, which are methods adapted to deal with modest space and time resources. This ties in with the relationship btw intensity-of-consciousness and degree-of-intelligence. In real life, these seem often to be tied together, because the cognitive structures that correlate with intensity of consciousness are useful ones for achieving intelligent behaviors. However, Searle's example is pathological in the sense that it posits a system with a high degree of intelligence associated with a functionality that is NOT associated with any intensity-of-consciousness. But I suggest that this pathology is due to the unrealistically large amount of computing resources that the rulebook requires. I.e., it is finitude of resources that causes intelligence and intensity-of-consciousness to be correlated. The fact that this correlation breaks in a pathological, physically-impossible case that requires dramatically much resources, doesn't mean too much... Note that I write about intensity of consciousness rather than presence of consciousness. I tend toward panpsychism but I do accept that while all animals are conscious, some animals are more conscious than others (to pervert Orwell). I have elaborated on this perspective considerably in The Hidden Pattern. -- Ben G - This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?member_id=231415user_secret=fabd7936
RE: [agi] Intelligence vs Efficient Intelligence
I'm probably not answering your question but have been thinking more on all this. There's the usual thermodynamics stuff and relativistic physics that is going on with intelligence and flipping bits within this universe, verses the no-friction universe or Newtonian setup. But what I've been thinking and this is probably just reiterating what someone else has worked through but basically a large part of intelligence is chaos control, chaos feedback loops, operating within complexity. Intelligence is some sort of delicate multi-vectored balancing act between complexity and projecting, manipulating, storing/modeling, NN training, genetic learning of the chaos and applying chaos in an environment and optimizing it's understanding and application of. The more intelligent, the better handle an entity has on the chaos. An intelligent entity can have maximal effect with minimal energy expenditure on its environment in a controlled manner; intelligence (or the application of) or even perhaps consciousness is the real-time surfing of buttery effects. So efficient intelligence involves thermodynamic power differentials of resource consumption applied to goals, etc. A goal would be expressed similarly to intelligence formulae. Really efficient means good chaos leverage understanding cycles, systems, entropy goings on over time and maximizing effect with minimal I/O control for goal achievement while utilizing the KR and the entity's resources... John From: Matt Mahoney [mailto:[EMAIL PROTECTED] I guess people want intelligence to be useful, not just complex :-) This raises a question. Suppose you had a very large program consisting of random instructions. Such a thing would have high algorithmic complexity, but most people would not say that such a thing was intelligent (depending on their favorite definition). But how would you know? If you didn't know how the code was generated, then how would you know that the program was really random and didn't actually solve some very hard class of problems? - This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?member_id=231415user_secret=fabd7936
Re: [agi] Relationship btw consciousness and intelligence
I liked most of your points, but . . . . However, Searle's example is pathological in the sense that it posits a system with a high degree of intelligence associated with a functionality that is NOT associated with any intensity-of-consciousness. But I suggest that this pathology is due to the unrealistically large amount of computing resources that the rulebook requires. Not by my definition of intelligence (which requires learning/adaptation). - Original Message - From: Benjamin Goertzel To: agi@v2.listbox.com Sent: Sunday, May 20, 2007 1:24 PM Subject: [agi] Relationship btw consciousness and intelligence Hi all, Someone emailed me recently about Searle's Chinese Room argument, http://en.wikipedia.org/wiki/Chinese_room a topic that normally bores me to tears, but it occurred to me that part of my reply might be of interest to some on this list, because it pertains to the more general issue of the relationship btw consciousness and intelligence. It also ties in with the importance of thinking about efficient intelligence rather than just raw intelligence, as discussed in the recent thread on definitions of intelligence. Here is the relevant part of my reply about Searle: However, a key point is: The scenario Searle describes is likely not physically possible, due to the unrealistically large size of the rulebook. The structures that we associate with intelligence (will, focused awareness, etc.) in a human context, all come out of the need to do intelligent processing within modest space and time requirements. So when we say we feel like the {Searle+rulebook} system isn't really understanding Chinese, what we mean is: It isn't understanding Chinese according to the methods we are used to, which are methods adapted to deal with modest space and time resources. This ties in with the relationship btw intensity-of-consciousness and degree-of-intelligence. In real life, these seem often to be tied together, because the cognitive structures that correlate with intensity of consciousness are useful ones for achieving intelligent behaviors. However, Searle's example is pathological in the sense that it posits a system with a high degree of intelligence associated with a functionality that is NOT associated with any intensity-of-consciousness. But I suggest that this pathology is due to the unrealistically large amount of computing resources that the rulebook requires. I.e., it is finitude of resources that causes intelligence and intensity-of-consciousness to be correlated. The fact that this correlation breaks in a pathological, physically-impossible case that requires dramatically much resources, doesn't mean too much... Note that I write about intensity of consciousness rather than presence of consciousness. I tend toward panpsychism but I do accept that while all animals are conscious, some animals are more conscious than others (to pervert Orwell). I have elaborated on this perspective considerably in The Hidden Pattern. -- Ben G -- This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?; - This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?member_id=231415user_secret=fabd7936
RE: [agi] Intelligence vs Efficient Intelligence
Oops heh I was eating French toast as I wrote this - intelligence (or the application of) or even perhaps consciousness is the real-time surfing of buttery effects I meant butterfly effects. John -Original Message- From: John G. Rose [mailto:[EMAIL PROTECTED] Sent: Sunday, May 20, 2007 11:45 AM To: agi@v2.listbox.com Subject: RE: [agi] Intelligence vs Efficient Intelligence I'm probably not answering your question but have been thinking more on all this. There's the usual thermodynamics stuff and relativistic physics that is going on with intelligence and flipping bits within this universe, verses the no-friction universe or Newtonian setup. But what I've been thinking and this is probably just reiterating what someone else has worked through but basically a large part of intelligence is chaos control, chaos feedback loops, operating within complexity. Intelligence is some sort of delicate multi-vectored balancing act between complexity and projecting, manipulating, storing/modeling, NN training, genetic learning of the chaos and applying chaos in an environment and optimizing it's understanding and application of. The more intelligent, the better handle an entity has on the chaos. An intelligent entity can have maximal effect with minimal energy expenditure on its environment in a controlled manner; intelligence (or the application of) or even perhaps consciousness is the real-time surfing of buttery effects. So efficient intelligence involves thermodynamic power differentials of resource consumption applied to goals, etc. A goal would be expressed similarly to intelligence formulae. Really efficient means good chaos leverage understanding cycles, systems, entropy goings on over time and maximizing effect with minimal I/O control for goal achievement while utilizing the KR and the entity's resources... John From: Matt Mahoney [mailto:[EMAIL PROTECTED] I guess people want intelligence to be useful, not just complex :-) This raises a question. Suppose you had a very large program consisting of random instructions. Such a thing would have high algorithmic complexity, but most people would not say that such a thing was intelligent (depending on their favorite definition). But how would you know? If you didn't know how the code was generated, then how would you know that the program was really random and didn't actually solve some very hard class of problems? - This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?; - This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?member_id=231415user_secret=fabd7936
Re: [agi] Relationship btw consciousness and intelligence
Sure... I prefer to define intelligence in terms of behavioral functionality rather than internal properties, but you are free to define it differently ;-) I note that if the Chinese language changes over time, then the {Searle + rulebook} system will rapidly become less intelligent in this context ben g On 5/20/07, Mark Waser [EMAIL PROTECTED] wrote: I liked most of your points, but . . . . However, Searle's example is pathological in the sense that it posits *a system with a high degree of intelligence* associated with a functionality that is NOT associated with any intensity-of-consciousness. But I suggest that this pathology is due to the unrealistically large amount of computing resources that the rulebook requires. Not by my definition of intelligence (which requires learning/adaptation). - Original Message - *From:* Benjamin Goertzel [EMAIL PROTECTED] *To:* agi@v2.listbox.com *Sent:* Sunday, May 20, 2007 1:24 PM *Subject:* [agi] Relationship btw consciousness and intelligence Hi all, Someone emailed me recently about Searle's Chinese Room argument, http://en.wikipedia.org/wiki/Chinese_room a topic that normally bores me to tears, but it occurred to me that part of my reply might be of interest to some on this list, because it pertains to the more general issue of the relationship btw consciousness and intelligence. It also ties in with the importance of thinking about efficient intelligence rather than just raw intelligence, as discussed in the recent thread on definitions of intelligence. Here is the relevant part of my reply about Searle: However, a key point is: The scenario Searle describes is likely not physically possible, due to the unrealistically large size of the rulebook. The structures that we associate with intelligence (will, focused awareness, etc.) in a human context, all come out of the need to do intelligent processing within modest space and time requirements. So when we say we feel like the {Searle+rulebook} system isn't really understanding Chinese, what we mean is: It isn't understanding Chinese according to the methods we are used to, which are methods adapted to deal with modest space and time resources. This ties in with the relationship btw intensity-of-consciousness and degree-of-intelligence. In real life, these seem often to be tied together, because the cognitive structures that correlate with intensity of consciousness are useful ones for achieving intelligent behaviors. However, Searle's example is pathological in the sense that it posits a system with a high degree of intelligence associated with a functionality that is NOT associated with any intensity-of-consciousness. But I suggest that this pathology is due to the unrealistically large amount of computing resources that the rulebook requires. I.e., it is finitude of resources that causes intelligence and intensity-of-consciousness to be correlated. The fact that this correlation breaks in a pathological, physically-impossible case that requires dramatically much resources, doesn't mean too much... Note that I write about intensity of consciousness rather than presence of consciousness. I tend toward panpsychism but I do accept that while all animals are conscious, some animals are more conscious than others (to pervert Orwell). I have elaborated on this perspective considerably in The Hidden Pattern. -- Ben G -- This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?; -- This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?; - This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?member_id=231415user_secret=fabd7936
Re: [agi] Write a doctoral dissertation, trigger a Singularity
Why is Murray allowed to remain on this mailing list, anyway? As a warning to others? The others don't appear to be taking the hint. -- Eliezer S. Yudkowsky http://singinst.org/ Research Fellow, Singularity Institute for Artificial Intelligence - This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?member_id=231415user_secret=fabd7936
Re: [agi] Write a doctoral dissertation, trigger a Singularity
Personally, I find many of his posts highly entertaining... If your sense of humor differs, you can always use the DEL key ;-) -- Ben G On 5/20/07, Eliezer S. Yudkowsky [EMAIL PROTECTED] wrote: Why is Murray allowed to remain on this mailing list, anyway? As a warning to others? The others don't appear to be taking the hint. -- Eliezer S. Yudkowsky http://singinst.org/ Research Fellow, Singularity Institute for Artificial Intelligence - This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?; - This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?member_id=231415user_secret=fabd7936
Re: [agi] Relationship btw consciousness and intelligence
Intelligence, to me, is the ability to achieve complex goals... This is one way of being functional a paperclip though is very functional yet not very intelligent... ben g On 5/20/07, Mark Waser [EMAIL PROTECTED] wrote: Sure... I prefer to define intelligence in terms of behavioral functionality rather than internal properties, but you are free to define it differently ;-) I wouldn't call learning/adaptability an internal(-only) property . . . . I note that if the Chinese language changes over time, then the {Searle + rulebook} system will rapidly become less intelligent in this context See. Now this indicates the funkiness of your definition . . . . Replace intelligent with functional and it makes a lot more sense. Actually, that raises a good question -- What is the difference between your intelligent and your functional? - Original Message - *From:* Benjamin Goertzel [EMAIL PROTECTED] *To:* agi@v2.listbox.com *Sent:* Sunday, May 20, 2007 2:11 PM *Subject:* Re: [agi] Relationship btw consciousness and intelligence Sure... I prefer to define intelligence in terms of behavioral functionality rather than internal properties, but you are free to define it differently ;-) I note that if the Chinese language changes over time, then the {Searle + rulebook} system will rapidly become less intelligent in this context ben g On 5/20/07, Mark Waser [EMAIL PROTECTED] wrote: I liked most of your points, but . . . . However, Searle's example is pathological in the sense that it posits *a system with a high degree of intelligence* associated with a functionality that is NOT associated with any intensity-of-consciousness. But I suggest that this pathology is due to the unrealistically large amount of computing resources that the rulebook requires. Not by my definition of intelligence (which requires learning/adaptation). - Original Message - *From:* Benjamin Goertzel [EMAIL PROTECTED] *To:* agi@v2.listbox.com *Sent:* Sunday, May 20, 2007 1:24 PM *Subject:* [agi] Relationship btw consciousness and intelligence Hi all, Someone emailed me recently about Searle's Chinese Room argument, http://en.wikipedia.org/wiki/Chinese_room a topic that normally bores me to tears, but it occurred to me that part of my reply might be of interest to some on this list, because it pertains to the more general issue of the relationship btw consciousness and intelligence. It also ties in with the importance of thinking about efficient intelligence rather than just raw intelligence, as discussed in the recent thread on definitions of intelligence. Here is the relevant part of my reply about Searle: However, a key point is: The scenario Searle describes is likely not physically possible, due to the unrealistically large size of the rulebook. The structures that we associate with intelligence (will, focused awareness, etc.) in a human context, all come out of the need to do intelligent processing within modest space and time requirements. So when we say we feel like the {Searle+rulebook} system isn't really understanding Chinese, what we mean is: It isn't understanding Chinese according to the methods we are used to, which are methods adapted to deal with modest space and time resources. This ties in with the relationship btw intensity-of-consciousness and degree-of-intelligence. In real life, these seem often to be tied together, because the cognitive structures that correlate with intensity of consciousness are useful ones for achieving intelligent behaviors. However, Searle's example is pathological in the sense that it posits a system with a high degree of intelligence associated with a functionality that is NOT associated with any intensity-of-consciousness. But I suggest that this pathology is due to the unrealistically large amount of computing resources that the rulebook requires. I.e., it is finitude of resources that causes intelligence and intensity-of-consciousness to be correlated. The fact that this correlation breaks in a pathological, physically-impossible case that requires dramatically much resources, doesn't mean too much... Note that I write about intensity of consciousness rather than presence of consciousness. I tend toward panpsychism but I do accept that while all animals are conscious, some animals are more conscious than others (to pervert Orwell). I have elaborated on this perspective considerably in The Hidden Pattern. -- Ben G -- This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?; -- This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?; -- This list is sponsored by AGIRI: http://www.agiri.org/email To
Re: [agi] Relationship btw consciousness and intelligence
Allow me to paraphrase . . . . Something is intelligent if it is functional over a wide variety of complex goals. Is that a reasonable shot at your definition? - Original Message - From: Benjamin Goertzel To: agi@v2.listbox.com Sent: Sunday, May 20, 2007 2:41 PM Subject: Re: [agi] Relationship btw consciousness and intelligence Intelligence, to me, is the ability to achieve complex goals... This is one way of being functional a paperclip though is very functional yet not very intelligent... ben g On 5/20/07, Mark Waser [EMAIL PROTECTED] wrote: Sure... I prefer to define intelligence in terms of behavioral functionality rather than internal properties, but you are free to define it differently ;-) I wouldn't call learning/adaptability an internal(-only) property . . . . I note that if the Chinese language changes over time, then the {Searle + rulebook} system will rapidly become less intelligent in this context See. Now this indicates the funkiness of your definition . . . . Replace intelligent with functional and it makes a lot more sense. Actually, that raises a good question -- What is the difference between your intelligent and your functional? - Original Message - From: Benjamin Goertzel To: agi@v2.listbox.com Sent: Sunday, May 20, 2007 2:11 PM Subject: Re: [agi] Relationship btw consciousness and intelligence Sure... I prefer to define intelligence in terms of behavioral functionality rather than internal properties, but you are free to define it differently ;-) I note that if the Chinese language changes over time, then the {Searle + rulebook} system will rapidly become less intelligent in this context ben g On 5/20/07, Mark Waser [EMAIL PROTECTED] wrote: I liked most of your points, but . . . . However, Searle's example is pathological in the sense that it posits a system with a high degree of intelligence associated with a functionality that is NOT associated with any intensity-of-consciousness. But I suggest that this pathology is due to the unrealistically large amount of computing resources that the rulebook requires. Not by my definition of intelligence (which requires learning/adaptation). - Original Message - From: Benjamin Goertzel To: agi@v2.listbox.com Sent: Sunday, May 20, 2007 1:24 PM Subject: [agi] Relationship btw consciousness and intelligence Hi all, Someone emailed me recently about Searle's Chinese Room argument, http://en.wikipedia.org/wiki/Chinese_room a topic that normally bores me to tears, but it occurred to me that part of my reply might be of interest to some on this list, because it pertains to the more general issue of the relationship btw consciousness and intelligence. It also ties in with the importance of thinking about efficient intelligence rather than just raw intelligence, as discussed in the recent thread on definitions of intelligence. Here is the relevant part of my reply about Searle: However, a key point is: The scenario Searle describes is likely not physically possible, due to the unrealistically large size of the rulebook. The structures that we associate with intelligence (will, focused awareness, etc.) in a human context, all come out of the need to do intelligent processing within modest space and time requirements. So when we say we feel like the {Searle+rulebook} system isn't really understanding Chinese, what we mean is: It isn't understanding Chinese according to the methods we are used to, which are methods adapted to deal with modest space and time resources. This ties in with the relationship btw intensity-of-consciousness and degree-of-intelligence. In real life, these seem often to be tied together, because the cognitive structures that correlate with intensity of consciousness are useful ones for achieving intelligent behaviors. However, Searle's example is pathological in the sense that it posits a system with a high degree of intelligence associated with a functionality that is NOT associated with any intensity-of-consciousness. But I suggest that this pathology is due to the unrealistically large amount of computing resources that the rulebook requires. I.e., it is finitude of resources that causes intelligence and intensity-of-consciousness to be correlated. The fact that this correlation breaks in a pathological, physically-impossible case that requires dramatically much resources, doesn't mean too much... Note that I write about intensity of consciousness rather than presence of consciousness. I tend toward panpsychism but I do accept that while all animals
Re: [agi] Relationship btw consciousness and intelligence
Sure, that's fine... I mean: I have given a mathematical definition before, so all these verbal paraphrases should be viewed as rough approximations anyway... On 5/20/07, Mark Waser [EMAIL PROTECTED] wrote: Allow me to paraphrase . . . . Something is intelligent if it is functional over a wide variety of complex goals. Is that a reasonable shot at your definition? - Original Message - *From:* Benjamin Goertzel [EMAIL PROTECTED] *To:* agi@v2.listbox.com *Sent:* Sunday, May 20, 2007 2:41 PM *Subject:* Re: [agi] Relationship btw consciousness and intelligence Intelligence, to me, is the ability to achieve complex goals... This is one way of being functional a paperclip though is very functional yet not very intelligent... ben g On 5/20/07, Mark Waser [EMAIL PROTECTED] wrote: Sure... I prefer to define intelligence in terms of behavioral functionality rather than internal properties, but you are free to define it differently ;-) I wouldn't call learning/adaptability an internal(-only) property . . . . I note that if the Chinese language changes over time, then the {Searle + rulebook} system will rapidly become less intelligent in this context See. Now this indicates the funkiness of your definition . . . . Replace intelligent with functional and it makes a lot more sense. Actually, that raises a good question -- What is the difference between your intelligent and your functional? - Original Message - *From:* Benjamin Goertzel [EMAIL PROTECTED] *To:* agi@v2.listbox.com *Sent:* Sunday, May 20, 2007 2:11 PM *Subject:* Re: [agi] Relationship btw consciousness and intelligence Sure... I prefer to define intelligence in terms of behavioral functionality rather than internal properties, but you are free to define it differently ;-) I note that if the Chinese language changes over time, then the {Searle + rulebook} system will rapidly become less intelligent in this context ben g On 5/20/07, Mark Waser [EMAIL PROTECTED] wrote: I liked most of your points, but . . . . However, Searle's example is pathological in the sense that it posits *a system with a high degree of intelligence* associated with a functionality that is NOT associated with any intensity-of-consciousness. But I suggest that this pathology is due to the unrealistically large amount of computing resources that the rulebook requires. Not by my definition of intelligence (which requires learning/adaptation). - Original Message - *From:* Benjamin Goertzel [EMAIL PROTECTED] *To:* agi@v2.listbox.com *Sent:* Sunday, May 20, 2007 1:24 PM *Subject:* [agi] Relationship btw consciousness and intelligence Hi all, Someone emailed me recently about Searle's Chinese Room argument, http://en.wikipedia.org/wiki/Chinese_room a topic that normally bores me to tears, but it occurred to me that part of my reply might be of interest to some on this list, because it pertains to the more general issue of the relationship btw consciousness and intelligence. It also ties in with the importance of thinking about efficient intelligence rather than just raw intelligence, as discussed in the recent thread on definitions of intelligence. Here is the relevant part of my reply about Searle: However, a key point is: The scenario Searle describes is likely not physically possible, due to the unrealistically large size of the rulebook. The structures that we associate with intelligence (will, focused awareness, etc.) in a human context, all come out of the need to do intelligent processing within modest space and time requirements. So when we say we feel like the {Searle+rulebook} system isn't really understanding Chinese, what we mean is: It isn't understanding Chinese according to the methods we are used to, which are methods adapted to deal with modest space and time resources. This ties in with the relationship btw intensity-of-consciousness and degree-of-intelligence. In real life, these seem often to be tied together, because the cognitive structures that correlate with intensity of consciousness are useful ones for achieving intelligent behaviors. However, Searle's example is pathological in the sense that it posits a system with a high degree of intelligence associated with a functionality that is NOT associated with any intensity-of-consciousness. But I suggest that this pathology is due to the unrealistically large amount of computing resources that the rulebook requires. I.e., it is finitude of resources that causes intelligence and intensity-of-consciousness to be correlated. The fact that this correlation breaks in a pathological, physically-impossible case that requires dramatically much resources, doesn't mean too much... Note that I write about intensity of consciousness rather than presence of consciousness. I tend toward
Re: [agi] Relationship btw consciousness and intelligence
Rough approximations maybe . . . . but you yourself have now pointed out that your definition is vulnerable to Searle's pathology (which is even simpler than the infinite AIXI effect :-) - Original Message - From: Benjamin Goertzel To: agi@v2.listbox.com Sent: Sunday, May 20, 2007 3:00 PM Subject: Re: [agi] Relationship btw consciousness and intelligence Sure, that's fine... I mean: I have given a mathematical definition before, so all these verbal paraphrases should be viewed as rough approximations anyway... On 5/20/07, Mark Waser [EMAIL PROTECTED] wrote: Allow me to paraphrase . . . . Something is intelligent if it is functional over a wide variety of complex goals. Is that a reasonable shot at your definition? - Original Message - From: Benjamin Goertzel To: agi@v2.listbox.com Sent: Sunday, May 20, 2007 2:41 PM Subject: Re: [agi] Relationship btw consciousness and intelligence Intelligence, to me, is the ability to achieve complex goals... This is one way of being functional a paperclip though is very functional yet not very intelligent... ben g On 5/20/07, Mark Waser [EMAIL PROTECTED] wrote: Sure... I prefer to define intelligence in terms of behavioral functionality rather than internal properties, but you are free to define it differently ;-) I wouldn't call learning/adaptability an internal(-only) property . . . . I note that if the Chinese language changes over time, then the {Searle + rulebook} system will rapidly become less intelligent in this context See. Now this indicates the funkiness of your definition . . . . Replace intelligent with functional and it makes a lot more sense. Actually, that raises a good question -- What is the difference between your intelligent and your functional? - Original Message - From: Benjamin Goertzel To: agi@v2.listbox.com Sent: Sunday, May 20, 2007 2:11 PM Subject: Re: [agi] Relationship btw consciousness and intelligence Sure... I prefer to define intelligence in terms of behavioral functionality rather than internal properties, but you are free to define it differently ;-) I note that if the Chinese language changes over time, then the {Searle + rulebook} system will rapidly become less intelligent in this context ben g On 5/20/07, Mark Waser [EMAIL PROTECTED] wrote: I liked most of your points, but . . . . However, Searle's example is pathological in the sense that it posits a system with a high degree of intelligence associated with a functionality that is NOT associated with any intensity-of-consciousness. But I suggest that this pathology is due to the unrealistically large amount of computing resources that the rulebook requires. Not by my definition of intelligence (which requires learning/adaptation). - Original Message - From: Benjamin Goertzel To: agi@v2.listbox.com Sent: Sunday, May 20, 2007 1:24 PM Subject: [agi] Relationship btw consciousness and intelligence Hi all, Someone emailed me recently about Searle's Chinese Room argument, http://en.wikipedia.org/wiki/Chinese_room a topic that normally bores me to tears, but it occurred to me that part of my reply might be of interest to some on this list, because it pertains to the more general issue of the relationship btw consciousness and intelligence. It also ties in with the importance of thinking about efficient intelligence rather than just raw intelligence, as discussed in the recent thread on definitions of intelligence. Here is the relevant part of my reply about Searle: However, a key point is: The scenario Searle describes is likely not physically possible, due to the unrealistically large size of the rulebook. The structures that we associate with intelligence (will, focused awareness, etc.) in a human context, all come out of the need to do intelligent processing within modest space and time requirements. So when we say we feel like the {Searle+rulebook} system isn't really understanding Chinese, what we mean is: It isn't understanding Chinese according to the methods we are used to, which are methods adapted to deal with modest space and time resources. This ties in with the relationship btw intensity-of-consciousness and degree-of-intelligence. In real life, these seem often to be tied together, because the cognitive structures that correlate with intensity of consciousness are useful ones for achieving intelligent behaviors.
Re: [agi] Relationship btw consciousness and intelligence
But I don't see vulnerability to Searle's pathology as a flaw in my definition of intelligence... The system {Searle + rulebook} **is** intelligent but not efficiently intelligent I conjecture that highly efficiently intelligent systems will necessarily possess intense consciousness and self-understanding. (Because I think that intense consciousness and self-understanding result from certain cognitive structures and dynamics, that I think are necessary for achieving efficient intelligence.) I don't think that high intelligence in principle implies intense consciousness or self-understanding... The reason this confuses people is that intelligence {roughly =} efficient intelligence for any real systems we have ever seen or know how to construct. The only intelligent but not efficiently intelligent systems we can talk about are hypothetical ones like {Searle+rulebook} or AIXI or AIXItl ... -- Ben G On 5/20/07, Mark Waser [EMAIL PROTECTED] wrote: Rough approximations maybe . . . . but you yourself have now pointed out that your definition is vulnerable to Searle's pathology (which is even simpler than the infinite AIXI effect :-) - Original Message - *From:* Benjamin Goertzel [EMAIL PROTECTED] *To:* agi@v2.listbox.com *Sent:* Sunday, May 20, 2007 3:00 PM *Subject:* Re: [agi] Relationship btw consciousness and intelligence Sure, that's fine... I mean: I have given a mathematical definition before, so all these verbal paraphrases should be viewed as rough approximations anyway... On 5/20/07, Mark Waser [EMAIL PROTECTED] wrote: Allow me to paraphrase . . . . Something is intelligent if it is functional over a wide variety of complex goals. Is that a reasonable shot at your definition? - Original Message - *From:* Benjamin Goertzel [EMAIL PROTECTED] *To:* agi@v2.listbox.com *Sent:* Sunday, May 20, 2007 2:41 PM *Subject:* Re: [agi] Relationship btw consciousness and intelligence Intelligence, to me, is the ability to achieve complex goals... This is one way of being functional a paperclip though is very functional yet not very intelligent... ben g On 5/20/07, Mark Waser [EMAIL PROTECTED] wrote: Sure... I prefer to define intelligence in terms of behavioral functionality rather than internal properties, but you are free to define it differently ;-) I wouldn't call learning/adaptability an internal(-only) property . . . . I note that if the Chinese language changes over time, then the {Searle + rulebook} system will rapidly become less intelligent in this context See. Now this indicates the funkiness of your definition . . . . Replace intelligent with functional and it makes a lot more sense. Actually, that raises a good question -- What is the difference between your intelligent and your functional? - Original Message - *From:* Benjamin Goertzel [EMAIL PROTECTED] *To:* agi@v2.listbox.com *Sent:* Sunday, May 20, 2007 2:11 PM *Subject:* Re: [agi] Relationship btw consciousness and intelligence Sure... I prefer to define intelligence in terms of behavioral functionality rather than internal properties, but you are free to define it differently ;-) I note that if the Chinese language changes over time, then the {Searle + rulebook} system will rapidly become less intelligent in this context ben g On 5/20/07, Mark Waser [EMAIL PROTECTED] wrote: I liked most of your points, but . . . . However, Searle's example is pathological in the sense that it posits *a system with a high degree of intelligence* associated with a functionality that is NOT associated with any intensity-of-consciousness. But I suggest that this pathology is due to the unrealistically large amount of computing resources that the rulebook requires. Not by my definition of intelligence (which requires learning/adaptation). - Original Message - *From:* Benjamin Goertzel [EMAIL PROTECTED] *To:* agi@v2.listbox.com *Sent:* Sunday, May 20, 2007 1:24 PM *Subject:* [agi] Relationship btw consciousness and intelligence Hi all, Someone emailed me recently about Searle's Chinese Room argument, http://en.wikipedia.org/wiki/Chinese_room a topic that normally bores me to tears, but it occurred to me that part of my reply might be of interest to some on this list, because it pertains to the more general issue of the relationship btw consciousness and intelligence. It also ties in with the importance of thinking about efficient intelligence rather than just raw intelligence, as discussed in the recent thread on definitions of intelligence. Here is the relevant part of my reply about Searle: However, a key point is: The scenario Searle describes is likely not physically possible, due to the unrealistically large size of the rulebook. The structures that we associate with
Re: [agi] Write a doctoral dissertation, trigger a Singularity
On 5/20/07, Benjamin Goertzel [EMAIL PROTECTED] wrote: Personally, I find many of his posts highly entertaining... If your sense of humor differs, you can always use the DEL key ;-) -- Ben G I initially found it sad and disturbing, no, disturbed. Thanks to Mark I was able to see the humor in it and I've created the appropriate filter to direct future such posts to my humor file. - Jef - This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?member_id=231415user_secret=fabd7936
Re: [agi] Relationship btw consciousness and intelligence
Actually, I think this a mistake, because it misses the core reason why Searle's argument is wrong, and repeats the mistake that he made. (I think, btw, that this kind of situation, where people come up with reasons against the CR arument that are not actually applicable or relevant, is one of the reasons for the CR argument's longevity. What I mean is: I think you are in good company here, because so many people have come up with so many of these sorts of arguments). The core reason for the failure of the CR is that it posits a situation in which an intelligence is implemented on top of another intelligence: then Searle makes an appeal to our feelings about the consciousness feelings of the wrong consciousness in this duo (the low level one). Can't do that: the consciousness of the top level intelligence is the only one that is relevant. Of course, the problem is that such a situation (one intelligence on top of another) is an exceptional case that one cannot make intuitive appeals about. Searle can scream all he wants that it makes no sense that there could be two intelligences here, but that just means he is ignorant about what intelligence is: it is not my job to fix Searle's ignorance. The reason your argument is a mistake is that it also makes reference to the conscious awareness of the low-level intelligence (at least, that is what it appears to be doing). As such, you are talking about the wrong intelligence, so your remarks are not relevant. Meta comment: I too find the CR deeply boring, but alas, you brought it up, so I had to say something ;-) Richard Loosemore. Benjamin Goertzel wrote: Hi all, Someone emailed me recently about Searle's Chinese Room argument, http://en.wikipedia.org/wiki/Chinese_room a topic that normally bores me to tears, but it occurred to me that part of my reply might be of interest to some on this list, because it pertains to the more general issue of the relationship btw consciousness and intelligence. It also ties in with the importance of thinking about efficient intelligence rather than just raw intelligence, as discussed in the recent thread on definitions of intelligence. Here is the relevant part of my reply about Searle: However, a key point is: The scenario Searle describes is likely not physically possible, due to the unrealistically large size of the rulebook. The structures that we associate with intelligence (will, focused awareness, etc.) in a human context, all come out of the need to do intelligent processing within modest space and time requirements. So when we say we feel like the {Searle+rulebook} system isn't really understanding Chinese, what we mean is: It isn't understanding Chinese according to the methods we are used to, which are methods adapted to deal with modest space and time resources. This ties in with the relationship btw intensity-of-consciousness and degree-of-intelligence. In real life, these seem often to be tied together, because the cognitive structures that correlate with intensity of consciousness are useful ones for achieving intelligent behaviors. However, Searle's example is pathological in the sense that it posits a system with a high degree of intelligence associated with a functionality that is NOT associated with any intensity-of-consciousness. But I suggest that this pathology is due to the unrealistically large amount of computing resources that the rulebook requires. I.e., it is finitude of resources that causes intelligence and intensity-of-consciousness to be correlated. The fact that this correlation breaks in a pathological, physically-impossible case that requires dramatically much resources, doesn't mean too much... Note that I write about intensity of consciousness rather than presence of consciousness. I tend toward panpsychism but I do accept that while all animals are conscious, some animals are more conscious than others (to pervert Orwell). I have elaborated on this perspective considerably in The Hidden Pattern. -- Ben G - This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?member_id=231415user_secret=fabd7936
RE: [agi] Intelligence vs Efficient Intelligence
--- John G. Rose [EMAIL PROTECTED] wrote: But what I've been thinking and this is probably just reiterating what someone else has worked through but basically a large part of intelligence is chaos control, chaos feedback loops, operating within complexity. Intelligence is some sort of delicate multi-vectored balancing act between complexity and projecting, manipulating, storing/modeling, NN training, genetic learning of the chaos and applying chaos in an environment and optimizing it's understanding and application of. The more intelligent, the better handle an entity has on the chaos. An intelligent entity can have maximal effect with minimal energy expenditure on its environment in a controlled manner; intelligence (or the application of) or even perhaps consciousness is the real-time surfing of buttery effects. I think the ability to model a chaotic process depends not so much on intelligence (whatever that is) as it does on knowledge of the state of the environment. For example, a chaotic process such as x := 4x(1 - x) has a really simple model. Your ability to predict x after 1000 iterations depends only on knowing the current value of x to several hundred decimal places. It is this type of knowledge that limits our ability to predict (and therefore control) the weather. I think there is a different role for chaos theory. Richard Loosemore describes a system as intelligent if it is complex and adaptive. Shane Legg's definition of universal intelligence requires (I believe) complexity but not adaptability. From a practical perspective I don't think it matters because we don't know how to build useful, complex systems that are not adaptive. For example, large software projects (code + human programmers) are adaptive in the sense that you can make incremental changes to the code without completely breaking the system, just as we incrementally update DNA or neural connections. One counterexample is a mathematical description of a cryptographic system. Any change to the system renders any prior analysis of its security invalid. Such systems are necessarily brittle. Out of necessity, we build systems that have mathematical descriptions simple enough to analyze. Stuart Kaufmann [1] noted that complex systems such as DNA tend to evolve to the boundary between stability and chaos, e.g. a Lyapunov exponent near 1 (or its approximation in discrete systems). I believe this is because overly stable systems aren't very complex (can't solve hard problems) and overly chaotic systems aren't adaptive (too brittle). [1] Kauffman, Stuart A., “Antichaos and Adaptation”, Scientific American, Aug. 1991, p. 64. -- Matt Mahoney, [EMAIL PROTECTED] - This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?member_id=231415user_secret=fabd7936
Re: [agi] Intelligence vs Efficient Intelligence
Matt Mahoney wrote: I think there is a different role for chaos theory. Richard Loosemore describes a system as intelligent if it is complex and adaptive. NO, no no no no! I already denied this. Misunderstanding: I do not say that a system as intelligent if it is complex and adaptive. Complex Adaptive System is a near-synonym for complex system, that's all. Richard Loosemore. - This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?member_id=231415user_secret=fabd7936
Re: [agi] Intelligence vs Efficient Intelligence
--- Richard Loosemore [EMAIL PROTECTED] wrote: Matt Mahoney wrote: I think there is a different role for chaos theory. Richard Loosemore describes a system as intelligent if it is complex and adaptive. NO, no no no no! I already denied this. Misunderstanding: I do not say that a system as intelligent if it is complex and adaptive. Complex Adaptive System is a near-synonym for complex system, that's all. OK, so what is your definition of intelligence? -- Matt Mahoney, [EMAIL PROTECTED] - This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?member_id=231415user_secret=fabd7936
RE: [agi] Intelligence vs Efficient Intelligence
Well I'm going into conjecture area because my technical knowledge of some of these disciplines is weak, but I'll keep going just for grins. Take an example of an entity existing in a higher level of consciousness - a Buddha who has achieved enlightenment. What is going on there? Verses and ant who operates in a lower level of consciousness, and then the average Joe who is in a different level of consciousness. Could it be that they are existing in different orbits or sweet spots/equilibria regions within a spectrum of environmental chaotic relationships? Can the enlightened Buddha have vast awareness as seeing cause and effect/butterfly effect as small distances verses the ant who can't see the distance between most cause/effects, only the very tiny ones? An AGI could run in different orbits/levels, and then this would allow for AGI's with really high levels of consciousness with tiny knowledge bases or vice versa. A Google for example is a massive KB running in a very low orbit. There are probably limits to the highest levels/orbits of consciousness. Also the orbits may induce some sort of brittleness for entities running within them if the entity is forced to and can't adapt to running outside of their home orbit... Just some thoughts. John From: Matt Mahoney [mailto:[EMAIL PROTECTED] I think the ability to model a chaotic process depends not so much on intelligence (whatever that is) as it does on knowledge of the state of the environment. For example, a chaotic process such as x := 4x(1 - x) has a really simple model. Your ability to predict x after 1000 iterations depends only on knowing the current value of x to several hundred decimal places. It is this type of knowledge that limits our ability to predict (and therefore control) the weather. I think there is a different role for chaos theory. Richard Loosemore describes a system as intelligent if it is complex and adaptive. Shane Legg's definition of universal intelligence requires (I believe) complexity but not adaptability. From a practical perspective I don't think it matters because we don't know how to build useful, complex systems that are not adaptive. For example, large software projects (code + human programmers) are adaptive in the sense that you can make incremental changes to the code without completely breaking the system, just as we incrementally update DNA or neural connections. One counterexample is a mathematical description of a cryptographic system. Any change to the system renders any prior analysis of its security invalid. Such systems are necessarily brittle. Out of necessity, we build systems that have mathematical descriptions simple enough to analyze. Stuart Kaufmann [1] noted that complex systems such as DNA tend to evolve to the boundary between stability and chaos, e.g. a Lyapunov exponent near 1 (or its approximation in discrete systems). I believe this is because overly stable systems aren't very complex (can't solve hard problems) and overly chaotic systems aren't adaptive (too brittle). [1] Kauffman, Stuart A., Antichaos and Adaptation, Scientific American, Aug. 1991, p. 64. -- Matt Mahoney, [EMAIL PROTECTED] - This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?member_id=231415user_secret=fabd7936
Re: [agi] Relationship btw consciousness and intelligence
Ben, Let me try to be mathematical and behavioral, too. Assume we finally agree on a way to measure a system's problem-solving capability (over a wide variety of complex goals) with a numerical function F(t), with t as the time of the measurement. The system's resources cost is also measured by a numerical function C(t). You and Shane believe that the value of F(t) is also a measurement of intelligence. Furthermore, you suggest efficient intelligence to be F(t)/C(t), and arguing that it is more realistic and relevant than raw intelligence. You also think my definition of intelligence is roughly the same. But to me, in this situation intelligence is better measured by F'(t), that is, the derivative of the capability, or how much the capability of the system can change (usually increase), under a constant resources supply. I believe it is also close to what Mark said. All these three measurement makes sense and are related to the everyday meaning of the word intelligence, though they are very different. For a system without adaptation ability, both F(t) and F(t)/C(t) can be large, but F'(t) is zero --- this is conventional computer systems, in my mind. On the other hand, systems with large F'(t) have great potentials, though initially may not have much problem-solving capability --- this is AI systems, according to my definition. For practical applications, we surely want systems with both large F(t)/C(t) and large F'(t), and system with huge F(t) at the cost of a huge C(t), like AIXI, is unrealistic --- we all agree here, including Shane, so it is not the issue. The issue is: F(t)/C(t) and F'(t) are different (though not the opposite of each other). Pei On 5/20/07, Benjamin Goertzel [EMAIL PROTECTED] wrote: Sure, that's fine... I mean: I have given a mathematical definition before, so all these verbal paraphrases should be viewed as rough approximations anyway... On 5/20/07, Mark Waser [EMAIL PROTECTED] wrote: Allow me to paraphrase . . . . Something is intelligent if it is functional over a wide variety of complex goals. Is that a reasonable shot at your definition? - Original Message - From: Benjamin Goertzel To: agi@v2.listbox.com Sent: Sunday, May 20, 2007 2:41 PM Subject: Re: [agi] Relationship btw consciousness and intelligence Intelligence, to me, is the ability to achieve complex goals... This is one way of being functional a paperclip though is very functional yet not very intelligent... ben g On 5/20/07, Mark Waser [EMAIL PROTECTED] wrote: Sure... I prefer to define intelligence in terms of behavioral functionality rather than internal properties, but you are free to define it differently ;-) I wouldn't call learning/adaptability an internal(-only) property . . . . I note that if the Chinese language changes over time, then the {Searle + rulebook} system will rapidly become less intelligent in this context See. Now this indicates the funkiness of your definition . . . . Replace intelligent with functional and it makes a lot more sense. Actually, that raises a good question -- What is the difference between your intelligent and your functional? - Original Message - From: Benjamin Goertzel To: agi@v2.listbox.com Sent: Sunday, May 20, 2007 2:11 PM Subject: Re: [agi] Relationship btw consciousness and intelligence Sure... I prefer to define intelligence in terms of behavioral functionality rather than internal properties, but you are free to define it differently ;-) I note that if the Chinese language changes over time, then the {Searle + rulebook} system will rapidly become less intelligent in this context ben g On 5/20/07, Mark Waser [EMAIL PROTECTED] wrote: I liked most of your points, but . . . . However, Searle's example is pathological in the sense that it posits a system with a high degree of intelligence associated with a functionality that is NOT associated with any intensity-of-consciousness. But I suggest that this pathology is due to the unrealistically large amount of computing resources that the rulebook requires. Not by my definition of intelligence (which requires learning/adaptation). - Original Message - From: Benjamin Goertzel To: agi@v2.listbox.com Sent: Sunday, May 20, 2007 1:24 PM Subject: [agi] Relationship btw consciousness and intelligence Hi all, Someone emailed me recently about Searle's Chinese Room argument, http://en.wikipedia.org/wiki/Chinese_room a topic that normally bores me to tears, but it occurred to me that part of my reply might be of interest to some on this list, because it pertains to the more general issue of the relationship btw consciousness and intelligence. It also ties in with the importance of thinking about efficient intelligence rather than just raw intelligence, as discussed in the recent thread on definitions of
Re: [agi] Relationship btw consciousness and intelligence
On 5/20/07, Mark Waser [EMAIL PROTECTED] wrote: Seems to me like you're going through *a lot* of effort for the same effect + a lot of confusion You conjecture that highly efficiently intelligent systems will necessarily possess intense consciousness and self-understanding. Isn't possess intense consciousness and self-understanding exactly the same as learn? They are not the same thing, although both are apparently necessary to achieve efficient intelligence So aren't you just saying that highly efficiently intelligent systems will necessarily learn? And why don't we just simplify highly efficiently intelligent as intelligent -- and just, by fiat, declare that anything that isn't highly efficiently intelligent is merely (at best) reflexively functional. You're just expressing a different taste in mapping formal definitions onto English phrases. I really don't think it matters what mapping you choose, so long as the mapping is defined clearly... However, I am coming to the opinion that mapping any formal definition into the NL term intelligence is a political error... From now on maybe I will use raw intelligence = complexity of goals achievable efficient intelligence = sum of (goal complexity)/(resources required to achieve goal) and not try to attach any single formal definition to the obviously highly ambiguous NL term intelligence .. -- Ben G -- Ben - This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?member_id=231415user_secret=fabd7936
RE: [agi] Intelligence vs Efficient Intelligence
--- John G. Rose [EMAIL PROTECTED] wrote: Well I'm going into conjecture area because my technical knowledge of some of these disciplines is weak, but I'll keep going just for grins. Take an example of an entity existing in a higher level of consciousness - a Buddha who has achieved enlightenment. What is going on there? Verses and ant who operates in a lower level of consciousness, and then the average Joe who is in a different level of consciousness. Could it be that they are existing in different orbits or sweet spots/equilibria regions within a spectrum of environmental chaotic relationships? Can the enlightened Buddha have vast awareness as seeing cause and effect/butterfly effect as small distances verses the ant who can't see the distance between most cause/effects, only the very tiny ones? An AGI could run in different orbits/levels, and then this would allow for AGI's with really high levels of consciousness with tiny knowledge bases or vice versa. A Google for example is a massive KB running in a very low orbit. There are probably limits to the highest levels/orbits of consciousness. Also the orbits may induce some sort of brittleness for entities running within them if the entity is forced to and can't adapt to running outside of their home orbit... I thought that Buddhist enlightenment meant realizing that seeking earthly pleasures (short term goals) is counterproductive to the longer term goal of happiness through enlightenment. Thus, the Buddha is more intelligent, if you measure intelligence by the ability to achieve goals. (But, being unenlightened, I could be wrong). But I don't see how you can measure the intelligence or consciousness of an attractor in a dynamical system. Also, I don't believe that consciousness is something that can be detected or measured. It is not a requirement for intelligence. What humans actually have is a belief in their own consciousness. It is part of your motivational system and cannot be changed. You could not disbelieve in your own consciousness or free will, even if you wanted to. -- Matt Mahoney, [EMAIL PROTECTED] - This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?member_id=231415user_secret=fabd7936
Re: [agi] Relationship btw consciousness and intelligence
On 5/20/07, Pei Wang [EMAIL PROTECTED] wrote: OK, it sounds much better than your previous descriptions to me (though there are still issues which I'd rather not discuss now). Much of our disagreement seems just to be about what goes in the def'n of intelligence and what goes in theorems about the properties required by intelligence. Which then largely becomes a matter of taste. But how about systems that cannot learn at all but have strong built-in capability and efficiency (within certain domains)? Will you say that they are intelligent but not too much, or not intelligent at all? I would say that they do have intelligence. But I would conjecture that there are strict limits to how much efficient intelligence such systems can have. -- Ben - This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?member_id=231415user_secret=fabd7936
Re: [agi] Relationship btw consciousness and intelligence
Adding onto the catalogue of specific sub-concepts of intelligence, we can identify not only raw intelligence = goal-achieving power efficient intelligence = goal-achieving power per unit of computational resources adaptive intelligence = ability to achieve goals newly presented to the system, not known to the system or its creators at the time of its creation [the wording could probably be improved] Shane wants to define intelligence as what I here call raw intelligence Pei and Mark Waser want to define intelligence as what I here call adaptive intelligence What is interesting to me is not which one of these various people want to identify with the NL term intelligence, but rather the relationships between the different types of intelligence. For example, many of us seem to support the conjecture that adaptive intelligence is necessary for efficient intelligence -- Ben G On 5/20/07, Benjamin Goertzel [EMAIL PROTECTED] wrote: On 5/20/07, Pei Wang [EMAIL PROTECTED] wrote: OK, it sounds much better than your previous descriptions to me (though there are still issues which I'd rather not discuss now). Much of our disagreement seems just to be about what goes in the def'n of intelligence and what goes in theorems about the properties required by intelligence. Which then largely becomes a matter of taste. But how about systems that cannot learn at all but have strong built-in capability and efficiency (within certain domains)? Will you say that they are intelligent but not too much, or not intelligent at all? I would say that they do have intelligence. But I would conjecture that there are strict limits to how much efficient intelligence such systems can have. -- Ben - This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?member_id=231415user_secret=fabd7936
Re: [agi] Relationship btw consciousness and intelligence
The reason your argument is a mistake is that it also makes reference to the conscious awareness of the low-level intelligence (at least, that is what it appears to be doing). As such, you are talking about the wrong intelligence, so your remarks are not relevant. I didn't mean to be doing that. Of course Searle, in the parable, has intensive consciousness, as well as efficient and adaptive intelligence But the knowledge of Chinese is immanent in the system {Searle+rulebook}, which does not have intensive consciousness, and in the context of knowing Chinese, has only raw intelligence but neither efficient nor highly adaptive intelligence... -- Ben G - This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?member_id=231415user_secret=fabd7936
Re: [agi] Relationship btw consciousness and intelligence
On 5/20/07, Benjamin Goertzel [EMAIL PROTECTED] wrote: Much of our disagreement seems just to be about what goes in the def'n of intelligence and what goes in theorems about the properties required by intelligence. Which then largely becomes a matter of taste. Part of them, yes, but not all --- our differences usually turn out to be larger than your estimation, though smaller than mine. :) But how about systems that cannot learn at all but have strong built-in capability and efficiency (within certain domains)? Will you say that they are intelligent but not too much, or not intelligent at all? I would say that they do have intelligence. You see, we do disagree here. To me, system that don't learn at all or assume infinite resources have zero intelligence, though they can be useful for other purposes. Pei - This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?member_id=231415user_secret=fabd7936
Re: [agi] Relationship btw consciousness and intelligence
On 5/20/07, Benjamin Goertzel [EMAIL PROTECTED] wrote: Adding onto the catalogue of specific sub-concepts of intelligence, we can identify not only raw intelligence = goal-achieving power efficient intelligence = goal-achieving power per unit of computational resources adaptive intelligence = ability to achieve goals newly presented to the system, not known to the system or its creators at the time of its creation [the wording could probably be improved] Again, it sounds much better, though adaptive intelligence sounds redundant to me. What is interesting to me is not which one of these various people want to identify with the NL term intelligence, but rather the relationships between the different types of intelligence. Agree. At least people should recognize them as different, and stop using one standard to evaluate another. Pei - This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?member_id=231415user_secret=fabd7936
Re: [agi] Relationship btw consciousness and intelligence
OK, it sounds much better than your previous descriptions to me (though there are still issues which I'd rather not discuss now). But how about systems that cannot learn at all but have strong built-in capability and efficiency (within certain domains)? Will you say that they are intelligent but not too much, or not intelligent at all? Pei On 5/20/07, Benjamin Goertzel [EMAIL PROTECTED] wrote: Actually, rather than F(t) we need to conceive capability as F(I) where I is a time interval... Then F(I) is the ability of the system to achieve complex goals over the interval I If I is long enough, then this encompasses the system's capability to learn to achieve new goals based on instruction, during the time interval I So I would say that having a high F'(I) over short time intervals I, is the right way to have a high (F/C)(I) over long time intervals I This basically is the conjecture that **learning is the path to efficient intelligence** -- Ben On 5/20/07, Pei Wang [EMAIL PROTECTED] wrote: Ben, Let me try to be mathematical and behavioral, too. Assume we finally agree on a way to measure a system's problem-solving capability (over a wide variety of complex goals) with a numerical function F(t), with t as the time of the measurement. The system's resources cost is also measured by a numerical function C(t). You and Shane believe that the value of F(t) is also a measurement of intelligence. Furthermore, you suggest efficient intelligence to be F(t)/C(t), and arguing that it is more realistic and relevant than raw intelligence. You also think my definition of intelligence is roughly the same. But to me, in this situation intelligence is better measured by F'(t), that is, the derivative of the capability, or how much the capability of the system can change (usually increase), under a constant resources supply. I believe it is also close to what Mark said. All these three measurement makes sense and are related to the everyday meaning of the word intelligence, though they are very different. For a system without adaptation ability, both F(t) and F(t)/C(t) can be large, but F'(t) is zero --- this is conventional computer systems, in my mind. On the other hand, systems with large F'(t) have great potentials, though initially may not have much problem-solving capability --- this is AI systems, according to my definition. For practical applications, we surely want systems with both large F(t)/C(t) and large F'(t), and system with huge F(t) at the cost of a huge C(t), like AIXI, is unrealistic --- we all agree here, including Shane, so it is not the issue. The issue is: F(t)/C(t) and F'(t) are different (though not the opposite of each other). Pei On 5/20/07, Benjamin Goertzel [EMAIL PROTECTED] wrote: Sure, that's fine... I mean: I have given a mathematical definition before, so all these verbal paraphrases should be viewed as rough approximations anyway... On 5/20/07, Mark Waser [EMAIL PROTECTED] wrote: Allow me to paraphrase . . . . Something is intelligent if it is functional over a wide variety of complex goals. Is that a reasonable shot at your definition? - Original Message - From: Benjamin Goertzel To: agi@v2.listbox.com Sent: Sunday, May 20, 2007 2:41 PM Subject: Re: [agi] Relationship btw consciousness and intelligence Intelligence, to me, is the ability to achieve complex goals... This is one way of being functional a paperclip though is very functional yet not very intelligent... ben g On 5/20/07, Mark Waser [EMAIL PROTECTED] wrote: Sure... I prefer to define intelligence in terms of behavioral functionality rather than internal properties, but you are free to define it differently ;-) I wouldn't call learning/adaptability an internal(-only) property . . . . I note that if the Chinese language changes over time, then the {Searle + rulebook} system will rapidly become less intelligent in this context See. Now this indicates the funkiness of your definition . . . . Replace intelligent with functional and it makes a lot more sense. Actually, that raises a good question -- What is the difference between your intelligent and your functional? - Original Message - From: Benjamin Goertzel To: agi@v2.listbox.com Sent: Sunday, May 20, 2007 2:11 PM Subject: Re: [agi] Relationship btw consciousness and intelligence Sure... I prefer to define intelligence in terms of behavioral functionality rather than internal properties, but you are free to define it differently ;-) I note that if the Chinese language changes over time, then the {Searle + rulebook} system will rapidly become less intelligent in this context ben g On 5/20/07, Mark Waser [EMAIL PROTECTED] wrote: I liked