"Most people define creativity as an inductive process that arrives at desirable results"
This is totally perverse. Try and find any mention of "induction" here (or in any text on creativity) http://en.wikipedia.org/wiki/Creativity The great majority of psychological and philosophical texts on creativity see logic as OPPOSED to creativity, which indeed it is. Creativity is any new course of action - involving any new,freeform mix of actions - to achieve a goal. It typically involves some "innovation" - coming up with new actions and objects. A new combination of steps, handling motions, placings of objects in a structure... You walk across a rock causeway, you're taking a creative walk, because you'll be taking a new combination of steps, and you'll have to come up with new steps/postures/balancings of your body. This can't be preplanned. And many robotic projects are trying to test themselves on walking across proto-causeways of rocks. The confusion re creativity arises because it has been largely confined to major, cultural, transformational creativity as distinct from everyday, personal, incremental creativity (wh. is what I just defined). In the former, you would have to not just produce a new mix of actions/steps, but devise a whole new kind of walking - to change the fundamental framework of the activity. A new *structural*/transformational mix.Invent a new kind of human kangaroo hop, say. No algo has ever introduced a single new element - if it did it would automatically be an AGI. An algo can say build a set of lego brick houses, but it can't introduce a single new building block/"brick". It can't suddenly without having been preprogrammed, introduce rocks, cardboard boxes, toys, corrugated iron sheets .et al... But humans can - because humans are endlessly creative in the simple incremental sense. Humans endlessly build new structures. Similarly humans endlessly walk on new terrains - where an algorithmic robot can only work on an old specially programmed terrain. All algos are superspecialist - that's why we talk about GENERAL AI , because no algo has ever been general, or other than hyperspecialist - has ever worked with more than one set of materials, or set of terrains, or set of conversational subjects. You, a human, can talk about endless new subjects, build endless new structures, travel endless new territories, You endlessly diversify in all your activities. IOnce an agent can incorporate just one new element - one new building material - then it can automatically endlessly incorporate new elements/bricks. Pretty well everything you do - every walk, conversation, reading of a book, viewing of a movie, playing of a field sport, tidying of a room - every single activity - is creative. Incorporates new actions on new objects. Nothing but nothing algos do is creative. Neither you nor anyone else has been or ever will be able to give **one fucking example** of an algo producing a new element. Just waffling on about how an algo invented x or y in the vaguest of terms, without specific examples of new elements, is a waste of breath. Algos only produce OLD courses of action, not new ones. Where you get confused is by using a trivial sense of "new". If a calculator multiplies 2365*56788 it may be "new" inasmuch as it has never multiplied that sum before. If you want to call that new, it's "new and formulaic" as opposed to "new and different" - just a predesigned permutation on the same old set of actions and objects, as distinct from a new and different, ad hooc incoporation of truly new actions.. God knows how many AI-ers think their chess programs are "creative" in playing "new" games. If a chess program could incorporate totally new pieces - merge,say, with snakes and ladders, or battlefield games, as you could - then it would be truly creative. Creativity is what you do when you when you procreate - mix your genes with new genes, not, like algos, (incl. genetic algos) the same old ones. On 9 October 2013 19:10, Steve Richfield <[email protected]> wrote: > Matt, > > On Wed, Oct 9, 2013 at 10:27 AM, Matt Mahoney <[email protected]>wrote: > >> On Wed, Oct 9, 2013 at 12:13 PM, tintner michael >> <[email protected]> wrote: >> > >> > Computers running algorithms can't be creative. Period. Never have >> never will produce one single new element - one new action, one new object. >> > > You must have a perverse definition of "creativity" in mind to make this > statement. What is it? > > I hold a U.S. patent on a bearing whose operating principle was discovered > by a computer. No, I did NOT declare the computer as co-inventor, and soon > thereafter scrapped the computer. The bearing utilized a then new principle > of automatically converting from liquid to gas lubrication when the shaft > speed become high enough, with the only moving parts being the shaft and > the liquid. This was stumbled into by my computer program, that was simply > trying random things and displaying the best results. > > Most people define creativity as an inductive process that arrives at > desirable results. Computers are quite good at induction, while many people > are quite poor at induction. Some game playing programs arrive at quite > creative solutions. > > Of course this required me to write the program, and to RECOGNIZE the > creative result. Like Schrodinger's Cat, creativity is in the mind of the > beholder and NOT in the process itself. > >> >> It's not that they can't be creative, but that we don't want them to >> be. > > > Sure we do. I used computers at Boeing to design some of the parts of > 737s, and then later flew in what was in part a computer's creation. > > Simple optimization is NOT enough to design in our modern era - which is > why Genetic Algorithms were developed. GAs often outperform human induction. > > We usually program computers to do the drudge work and save the >> fun stuff for ourselves. >> > > ... like being able to point our fingers at something and say "THAT is > creative. Oh, it was done by a computer? Then, I guess it isn't really > creative." > >> >> But creativity doesn't require a lot of advanced technology. >> http://www.youtube.com/watch?v=X4Neivqp2K4 >> > > ... and before that, a computer to solve complex 5x5 non-zero-sum war > strategy problems that would stump almost any human. My computer even > earned an award from USAF General Curtis LeMay. Attached is a photo of me > demonstrating my computer at WESCON in 1962. > > Steve > > *AGI* | Archives <https://www.listbox.com/member/archive/303/=now> > <https://www.listbox.com/member/archive/rss/303/6952829-59a2eca5> | > Modify<https://www.listbox.com/member/?&;>Your Subscription > <http://www.listbox.com> > ------------------------------------------- AGI Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/21088071-f452e424 Modify Your Subscription: https://www.listbox.com/member/?member_id=21088071&id_secret=21088071-58d57657 Powered by Listbox: http://www.listbox.com
