Matt, If your program can't handle natural language with all its ambiguities, > then it isn't AGI.
Internet AGIs are the technology of the future, and always will be. There will NEVER EVER in a million years be a thinking Internet silicon intelligence that will be able to solve substantial real-world problems based only on what exists on the Internet. I think that my prior email was pretty much a closed-form proof of that. However, there are MUCH simpler methods that work TODAY, given the metadata that is presently missing from the Internet. > >No one has yet proposed ANY way of "mining the Internet" to engage in > useful problem solving without these two pieces of metadata, yet supposedly > smart people continue wasting their efforts and other people's money on such > folly. > > My AGI proposal ( http://www.mattmahoney.net/agi2.html ) uses natural > language to communicate between peers. A peer is only required to understand > a small subset, perhaps scanning for a few keywords and ignoring everything > else. Individually, peers don't need to be very smart for the collective to > achieve AGI. While VERY interesting, your proposal appears to leave the following important questions unanswered: 1. How is it an AGI? I suppose this is a matter of definitions. It looks to me more like a protocol. 2. As I explained earlier on this thread, all human-human languages have severe semantic limitations, such that (applying this to your porposal), only very rarely will there ever exist an answer that PRECISELY answers a question, so some sort of "acceptable error" must go into the equation. In the example you used in your paper, Jupiter is NOT the largest planet that is known, as the astronomers have identified larger planets in other solar systems. There may be a good solution to this, e.g. provide the 3 best answers that are semantically disjoint. 3. Your paper addresses question answering, which as I have explained here in the past, is a much lower form of art than is problem solving, where you simply state an unsatisfactory situation and let the computer figure out why things are as they are and how to improve them. Note that http://www.DrEliza.com makes no attempt to answer questions, or even work on easy problems (because they are too damn hard for Dr. Eliza), but rather, it confines itself to very difficult corners of chosen sub-domains. It may never be able to match an average doctor, but can easily handle the fallout from the best of them. It will never match an average dentist, but it can save a lot of teeth that the best of them have given up on. Of course, your approach COULD be easily extended to function in problem solving, by simply providing a mechanism for users to attach the metadata that I have mentioned is needed for problem solving. In short, I really like your proposal for an alternative Internet protocol, as the Internet obviously needs a bunch of them, because the present set is woefully inadequate. IMHO you should recast your proposal as an RFC and put it out there. It sounds like you could easily utilize a USENET group for early demos. Note that Microsoft maintains some test groups on some of its servers, that Dr. Eliza already uses without problems for its inter-incarnation communication. Steve Richfield ------------------------------------------- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?member_id=8660244&id_secret=120640061-aded06 Powered by Listbox: http://www.listbox.com
