[agi] AGI Taxonomy
Is there a standard taxonomy of AGI that is referred to when talking about different AGIs or near AGIs? Saying that a software is an AGI or not an AGI is not descriptive enough. There are probably very few AGIs but many close AGIs and then many, many AIs. Software programs are like the plant and animal kingdom since they breed and multiply and evolve... John - 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] AGI Taxonomy
There's certainly no standard... at http://www.agiri.org/wiki/index.php?title=AGI_Projects I used 3 crude categories -- Neural net based -- Logic based -- Integrative ;-) On 5/9/07, John G. Rose [EMAIL PROTECTED] wrote: Is there a standard taxonomy of AGI that is referred to when talking about different AGIs or near AGIs? Saying that a software is an AGI or not an AGI is not descriptive enough. There are probably very few AGIs but many close AGIs and then many, many AIs. Software programs are like the plant and animal kingdom since they breed and multiply and evolve... John - 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] AGI Taxonomy
In Beyond AI I have a taxonomy (and Kurzweil picked that chapter, among others, to post on his site). in brief: Hypohuman AI -- below human ability and under human control Diahuman AI -- somewhere in the human range (which is large!) Epihuman AI -- smarter/more capable than human, but equivalent to a moderate-sized company (of very smart people) Hyperhuman AI -- equivalent to or better than all humans working in a given subject area and two involving a design stance rather than a capability level Parahuman AI -- designed to work alongside humans and relate Allohuman AI -- optimized for other things in ways that humans have a harder time relating to Josh On Wednesday 09 May 2007 06:14, John G. Rose wrote: Is there a standard taxonomy of AGI that is referred to when talking about different AGIs or near AGIs? Saying that a software is an AGI or not an AGI is not descriptive enough. There are probably very few AGIs but many close AGIs and then many, many AIs. Software programs are like the plant and animal kingdom since they breed and multiply and evolve... John - 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] AGI Taxonomy
My feeling is that it is better to classify the AGI projects alone multiple dimensions, rather than a single one. 1. Their exact goal (or their working definition of intelligence). On this aspect, I've tried to put them into 5 groups: * structure (e.g., to build brain model) * behavior (e.g., to simulate human mind) * capability (e.g., to solve hard problems) * function (e.g., to have cognitive facilities) * principle (e.g., to establish general theory) examples are in http://www.cis.temple.edu/~pwang/203-AI/Lecture/AGI.htm 2. Their technical strategy. So far I see 3 schools: * to integrate existing AI techniques (some people in mainstream AI are moving in this direction) * to establish an overall architecture, with modules that are based on different techniques (some mainstream AI people do this under the name of cognitive architecture; integrative AGI projects are also in this school) * to develop a unified core technique, then to extend it in various directions (some AGI projects mainly depend on a single technology) 3. Their major technology. This list is never complete, though the most common ones are: *. logic *. probability theory *. knowledge base *. production system *. natural language processing (the above are often collectively called symbolic) *. neural network *. evolutionary computation *. robotics Pei On 5/9/07, John G. Rose [EMAIL PROTECTED] wrote: Is there a standard taxonomy of AGI that is referred to when talking about different AGIs or near AGIs? Saying that a software is an AGI or not an AGI is not descriptive enough. There are probably very few AGIs but many close AGIs and then many, many AIs. Software programs are like the plant and animal kingdom since they breed and multiply and evolve... John - 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] AGI Taxonomy
I don't think intelligence can be measured that easily on a one dimensional axis, with a dot marking the intelligence of humans. If you look at all the possible intelligences, not just the organic ones we know of, measuring intelligence becomes extremely difficult. Measuring the intelligence of humans has been difficult in the past and so far we pretty much only have IQ (meaning logical tasks) test. Many have opposed this, saying that intelligence (in humans) is really something more than simple logic. Emotional intelligence,etc. If measuring the intelligence in VERY SIMILAR systems (compared to the vast 'mind design space') is difficult, then measuring all intelligences must be near impossible (unless someone can pull out that magic definition of intelligence). It also seems *very* human-centric to compare everything to humans.. Maybe measuring intelligence is like measuring how good a tool is. It depends on what you need it for. 2007/5/9, J. Storrs Hall, PhD. [EMAIL PROTECTED]: In Beyond AI I have a taxonomy (and Kurzweil picked that chapter, among others, to post on his site). in brief: Hypohuman AI -- below human ability and under human control Diahuman AI -- somewhere in the human range (which is large!) Epihuman AI -- smarter/more capable than human, but equivalent to a moderate-sized company (of very smart people) Hyperhuman AI -- equivalent to or better than all humans working in a given subject area and two involving a design stance rather than a capability level Parahuman AI -- designed to work alongside humans and relate Allohuman AI -- optimized for other things in ways that humans have a harder time relating to Josh On Wednesday 09 May 2007 06:14, John G. Rose wrote: Is there a standard taxonomy of AGI that is referred to when talking about different AGIs or near AGIs? Saying that a software is an AGI or not an AGI is not descriptive enough. There are probably very few AGIs but many close AGIs and then many, many AIs. Software programs are like the plant and animal kingdom since they breed and multiply and evolve... John - 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] Determinism
I work very hard to produce the exact same answer to the same question. If some humans don't actually do that, then they are just exhibiting the flaws that exist in our design. This is not to be confused with answering better over time, based on more and better information. The exact same information should always produce the exact same result in human or AGI. Irrational thought could be simulated by an AGI so that a better model of some humans could be had but the less intentional defects built into the AGI the better. A computer with finite memory can only model (predict) a computer with less memory. No computer can simulate itself. When we introspect on our own brains, we must simplify the model to a probabilistic one, whether or not it is actually deterministic. This is NOT true. How many answers can be had by the formula for a single straight line? The answer is infinite. A computer CAN model/simulate anything including itself (whatever that means) given enough time. If the model has understanding (formulas or algorithms) then any amount of simulated detail can be realized. David Clark - Original Message - From: Matt Mahoney [EMAIL PROTECTED] To: agi@v2.listbox.com Sent: Tuesday, May 08, 2007 12:47 PM Subject: Re: [agi] Determinism I really hate to get into this endless discussion. I think everyone agrees that some randomness in AGI decision making is good (e.g. learning through exploration). Also it does not matter if the source of randomness is a true random source, such as thermal noise in neurons, or a deterministic pseudo random number generator, such as iterating a cryptographic hash function with a secret seed. I think what is confusing Mike (and I am sure he will correct me) is that the inability of humans to predict their own thoughts (what will I later decide to have for dinner?) is something that needs to be programmed into an AGI. There is actually no other way to program it. A computer with finite memory can only model (predict) a computer with less memory. No computer can simulate itself. When we introspect on our own brains, we must simplify the model to a probabilistic one, whether or not it is actually deterministic. -- 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/?; - 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] AGI Taxonomy
Notice that I didn't use the word intelligence -- the key issue here is when we can expect the existence of AGI to make a significant difference in the world. Computers have had a big impact because they have abilities well beyond those of humans in certain limited areas. Of course, so did steam shovels. The key issue is ability, and it assumes a context that specifies what kind of ability we're talking about. For AGIs it would be those abilities that currently remain exclusive to humans. We live in a world whose parameters are largely couched in those terms -- for better or worse. That won't be true 50 years from now, for the first time in history. At that point it would make sense to come up with a new scale. Josh On Wednesday 09 May 2007 11:44, Panu Horsmalahti wrote: I don't think intelligence can be measured that easily on a one dimensional axis, with a dot marking the intelligence of humans. If you look at all the possible intelligences, not just the organic ones we know of, measuring intelligence becomes extremely difficult. Measuring the intelligence of humans has been difficult in the past and so far we pretty much only have IQ (meaning logical tasks) test. Many have opposed this, saying that intelligence (in humans) is really something more than simple logic. Emotional intelligence,etc. If measuring the intelligence in VERY SIMILAR systems (compared to the vast 'mind design space') is difficult, then measuring all intelligences must be near impossible (unless someone can pull out that magic definition of intelligence). It also seems *very* human-centric to compare everything to humans.. Maybe measuring intelligence is like measuring how good a tool is. It depends on what you need it for. 2007/5/9, J. Storrs Hall, PhD. [EMAIL PROTECTED]: In Beyond AI I have a taxonomy (and Kurzweil picked that chapter, among others, to post on his site). in brief: Hypohuman AI -- below human ability and under human control Diahuman AI -- somewhere in the human range (which is large!) Epihuman AI -- smarter/more capable than human, but equivalent to a moderate-sized company (of very smart people) Hyperhuman AI -- equivalent to or better than all humans working in a given subject area and two involving a design stance rather than a capability level Parahuman AI -- designed to work alongside humans and relate Allohuman AI -- optimized for other things in ways that humans have a harder time relating to Josh On Wednesday 09 May 2007 06:14, John G. Rose wrote: Is there a standard taxonomy of AGI that is referred to when talking about different AGIs or near AGIs? Saying that a software is an AGI or not an AGI is not descriptive enough. There are probably very few AGIs but many close AGIs and then many, many AIs. Software programs are like the plant and animal kingdom since they breed and multiply and evolve... John - 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/?; - 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] Determinism
--- David Clark [EMAIL PROTECTED] wrote: A computer with finite memory can only model (predict) a computer with less memory. No computer can simulate itself. When we introspect on our own brains, we must simplify the model to a probabilistic one, whether or not it is actually deterministic. This is NOT true. How many answers can be had by the formula for a single straight line? The answer is infinite. A computer CAN model/simulate anything including itself (whatever that means) given enough time. If the model has understanding (formulas or algorithms) then any amount of simulated detail can be realized. By simulate, I mean in the formal sense, as a universal Turing machine can simulate any other Turing machine, for example, you can write a program in C that runs programs written in Pascal (e.g. a compiler or interpreter). Thus, you can predict what the Pascal program will do. Languages like Pascal and C define Turing machines. They have unlimited memory. Real machines have finite memory, so you do the simulation properly you need to also define the hardware limits of the target machine. So if the real program reports an out of memory error, the simulation should too, at precisely the same point. Now if the target machine (running Pascal) has 2 MB memory, and your machine (running C) has 1 MB, then you can't do it. Your simulator will run out of memory first. Likewise, you can't simulate your own machine, because you need additional memory to run the simulator. When we lack the memory for an exact simulation, we can use an approximation, one that usually but not always gives the right answer. For example, we forecast the weather using an approximation of the state of the Earth's atmosphere and get an approximate answer. We can do the same with programs. For example, if a program outputs a string of bits according to some algorithm, then you can often predict most of the bits by looking up the last few bits of context in a table and predicting whatever bit was last output in this context. The cache and branch prediction logic in your CPU do something like this. This is an example of your computer simulating itself using a simplified, probabilistic model. A more accurate model would analyze the entire program and make exact predictions, but this is not only impractical but also impossible. So we must have some cache misses and branch mispredictions. In the same way, the brain cannot predict itself. The brain has finite memory. Even if the brain were deterministic (no neuron noise), this would still be the case. If a powerful enough computer knew the exact state of your brain, it could predict what you would think next, but you could not predict what that computer would output. I know in theory you could follow the computer's algorithm on pencil and paper, but even then you would still not know the result of that manual computation until you did it. No matter what you do, you cannot predict your own thoughts with 100% accuracy. Your mental model must be probabilistic, whether the hardware is deterministic or not. -- 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
[agi] Help get the 400k SIAI matching challenge on DIGG's front page
Hello, As you may know the SIAI has started a matching challenge of 400'000 USD please help to get the word out by digging the story and thereby putting it on Digg's front page: http://digg.com/general_sciences/SIAI_seeks_funding_for_AI_research Xie Xie, Stefan -- Stefan Pernar App. 1-6-I, Piao Home No. 19 Jiang Tai Xi Lu 100016 Beijing China Mobil: +86 1391 009 1931 Skype: Stefan.Pernar - 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