John Nilsson <j...@milsson.nu> writes: > I read that post about constraints and kept thinking that it should be > the infrastructure for the next generation of systems development, not > art assets :) > > In my mind it should be possible to input really fuzzy constraints > like "It should have a good looking, blog-like design" > A search engine would find a set of implications from that statement > created by designers and vetted by their peers. Some browsing and > light tweaking and there, I have a full front-end design provided for > the system. > > Then I add further constraints. "Available via http://blahblah.com/ > and be really cheap", again the search engine will find the implied > constrains and provide options among the cheaper cloud providers. I > pick one of them and there provisioning is taken care of. > > I guess the problem is to come up with a way to formalize all this > knowledge experts are sitting on into a representation usable by that > search engine. But could this not be done implicitly from the act of > selecting a match after a search? > > Say some solution S derived from constrains A,B,C is selected in my > search. I have constraint A,B and D as input. By implication the > system now knows that S is a solution to D.
Right. Just a "simple" applicaition of AI and all the algorithms developed so far. You just need to integrate them in a working system. And who has the resources to do this work: it seems to me to be a big endeavour. Collecting the research "prototype" developed during the last 50 years, and develop a such a product. Even Watson or Siri would only represent a small part of it. -- __Pascal Bourguignon__ http://www.informatimago.com/ A bad day in () is better than a good day in {}. _______________________________________________ fonc mailing list fonc@vpri.org http://vpri.org/mailman/listinfo/fonc