You still haven't given a problem example - and never will. You can neither start with a set nor as you imply end with - "learn" - a set, when you are dealing with real world problems.
You have been talking prior to this about sets - causets/posets et al - being crucial to how the brain or an AGI solves problems. They have nothing to do with real world problems. And absolutely nothing to do with any classic, major creative problemsolving - like creating programs/software, solving mathematical theoretical problems, creating anything in the arts, sciences or technology. You may create a set as a solution to a software problem, but no set solved the problem in the first place. (There would be no such thing as the "creative agony" if there were any truth to that). So next time you mention EI, remember it stands for Evidentially Impossible. If you did wish to discuss this, we can take any software creation problem you like, and you will find no set of any kind, however you care to twist or sophisticate the concept. Real world problems are like, and include, detective problems - there is never a formulaic set of suspects. From: Sergio Pissanetzky Sent: Wednesday, June 13, 2012 5:24 PM To: AGI Subject: RE: [agi] The 2 Tests of AGI - generalizability & creativity Mike, Who says sets are going to solve a problem? It's inference. You need inference to solve problems. Inference works on knowledge, that you collect in sets. Is this philosophy? I don't think so. And it is you who brings me to philosophy all the time. I want to build my machine. There is an infinite variety of problem statements written by SM engineers who want developers to write software to support their respective companies. I have myself worked in two different companies, big important companies, Southwestern Bell and Delta Airlines, and that was precisely the work I did, talking to SM engineers and the developers (SM engineers are precious to the company, so they are shielded from developers). I've seen my share of problem statements, but they are all confidential, company property. And I have seen the entire process where software is created and released, and the mountains of documentation that are generated. Each time you call a number in the SW Bell area you are using my work, because I worked with the switches. If you fly Delta or use a kiosk or ask a question from an agent you are using my work. Go tell me about creative. Ask someone who is in software, perhaps Ben can give you the documentation for OpenCog. I selected an example used in European universities to teach refactoring. It is a token ring, sort of an intra-net that supports many computers that communicate with each other, for example in a University campus. I published the whole thing, just go and read it, I don't have time to present my work to you, go and find it yourself. But that's not all. Just in case you were planning to tell me something nasty. It doesn't work. None of that works. It's a ridiculous make up that managers use to appear as if they were in control and draw money from big companies. Do you know how software is really created? A man creates software! Only a man can create software. A human, man or woman. All those program statements change all the time, all the software changes, people are driven crazy, the workplace is in chaos. There are books and more books, methods and more methods to supposedly put all that under control. They don't work. Everything changes, there is an infinite variety of combinations, it all becomes fluid, squishy, it evades any attempt at organization. Until a human takes charge. Only a human has creativity enough to create the code. And the code is delivered and deployed, and the following day . guess what . there are changes. Only now it is called maintenance. Again, only humans can do it. Demolish? Looking at actual problems? You think I don't know about real world problems? Ridiculous! Real world problems are what motivate my theory. Outline the set for a squirrell? Animals don't HAVE a set, they LEARN their set as they go and they are creative as they go, and their creativity comes from what they learn. Outlining a set is narrow AI, and that is precisely what I oppose. And what you oppose, it seems to me. Sergio Sorry, I have other things to do. Good night. From: Mike Tintner [mailto:[email protected]] Sent: Wednesday, June 13, 2012 9:28 AM To: AGI Subject: Re: [agi] The 2 Tests of AGI - generalizability & creativity Sergio, Please give an example not an analysis of what you consider a creative or real world problem - and how you use a set to solve it. For example,s.o. may ask you : "write a program that will enable me to compose video essays - an alternative to ibooks Author.." - that's a problem or by all means, give me the problem that your subject-matter engineer or whoever gave you, in the form given - the original brief or give any other real world problem as it is originally expressed/conceived ("Jump on a tree branch", pace your squirrel, "jump over that box", "putt that golf ball to hole it" are also real/ real world problems) And then explain what set will solve that problem. You say you're a scientist (although what you're arguing below is nothing but philosophy). A scientist looks at evidence to test his theories. You say sets will solve real world problems - show me one piece of evidence - one actual, clearly identified real world problem that bears this out.. ONe way or another, looking at actual problems will help your theory - even if only to demolish it. Not looking at them won't help it at all. P.S. If you want to outline the set for a squirrel or a human jumping on a branch - [perhaps the worst example you could have chosen] - be my guest. AGI | Archives | Modify Your Subscription ------------------------------------------- AGI Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/21088071-c97d2393 Modify Your Subscription: https://www.listbox.com/member/?member_id=21088071&id_secret=21088071-2484a968 Powered by Listbox: http://www.listbox.com
