First of all, I do not believe science can allways respect past exact definitions of words in order to make progress. How about if Einstein refrained from publishing his relativity theory, because it would contradict the way people normally think about time/gravity? Beeing conservative in this aspect is the same thing as writing a program, without ever going back and changing names, revising old code. In the end you would be left with a mess.
Secondly, the scientific field of AI has already proved their incompetence in producing anything near AGI. It is apparent the field is too much focused on narrow AI, so why should we respect their narrow definitions which probably are set up to lead us into a hole again? Basically: Screw them, they have had their chance... ;-). Thirdly, the definition you outline on the topic is so broad that actually all learning systems that maintain and improve a set of solutions would fit into it. "select some set of the population and make a new population" could basically mean anything. The designs I have considered involve maintaining a set of favoured model components, and combining them to produe new models. So because I use a different definition of evolution, compared to this more narrow form of evolution you refer to, your attack on the inneficiency of "narrow evolution" is not directly relevant to anything that I said previously. Also, I do not believe that the human mind creates optimal routes from point A to point B. /Robert W 2007/10/20, William Pearson <[EMAIL PROTECTED]>: > > On 20/10/2007, Robert Wensman <[EMAIL PROTECTED]> wrote: > > I am not exactly sure how GA(genetic algorithm) and GP (genetic > programming) > > is defined. It seems that the concept of gene and evolution are very > much > > interconnected, so how we define genetic algorithm and genetic > programming > > depends on how we define evolutionary learning, which is partially a > topic > > of this thread. > > There is an academic field dedicated to these systems, it will confuse > people that know this field if you define GA and GP differently to > that field. There is also a field called evolutionary programming, > just to confuse things further. > > A rough definition of these type of evolutionary system is a system > with this rough pseudo code. > > 1. Generate random population > 2. Evaluate population > 3. Select some of the population to make up the new population > 4. Vary the population in some fashion > 5. Go to 2. > > To get a system capable of intelligence, based on some notion of > evolution, I'd argue you would have to break this template. The > trouble is that for the system to change you need to evaluate the > population, or even a single member, and then perform a selection. > > This is slower than other ways of getting information about the world, > from the world. Consider a robot trying to find the quickest path from > one part of the school to another. You could try lots of paths, > mutate them etc, or you could deduce a path from a map, that you find > in the school. Random variation and selection is not fit for the heavy > lifting part of learning. > > > > So basically yes, making evolutionary learning work fast enough is what > AGI > > is all about. But I do not feel that these methods to try to speed > things up > > make it less of an evolution, at least not in my opinion. The reason I > like > > the concept of evolutionary learning is that it implies some form of > open > > endedness, similar to how we think the thoughts of an intellect can go > in > > any direction. The words learning and adaptation has been too much used > in > > narrow AI in over simplified contexts. > > I, too, like the open ended-ness of evolution, but you can take that > part of it and use it for what it is good for, and ignore the other > parts of Darwinian evolution (random variation, hard > genotype-phenotype separation) and get a much more powerful system and > also open ended in more ways. > > Imagine a standard PC altered as much as is needed to do the > following. Unlike a normal PC there is no "root" user or dictatorial > all powerful operating system. Which programs have access to which > resources is governed by how much 'utility' they have been rewarded > and how they spend it. Resources include memory/processing power, and > if the program runs out of those it no longer exists. So there is some > form of selection within the system, but it would be a slow process > possibly only occurring during down time. It is vastly outpaced by the > changes in the system due to the programs altering themselves or > getting new data from the environment. Variation can be anything from > downloading a program from the internet to something randomly > generated. It is not a smoothly adapting system as you tend to be > thinking of. Would you still call it an evolutionary system? > > Even if you would, I was more cautioning you that other people would > think of Darwinian evolution and the limitations of that, when they > were told you were interested in evolutionary systems. And might > dismiss your ideas prematurely because of their experiences with those > types of evolutionary systems. > > Will Pearson > > ----- > 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=8660244&id_secret=55814931-efe642
