Talk:Artificial Intelligence

From InfoAnarchy
Jump to: navigation, search

This was originally on the AI page, and i'm not sure how it's related. Was this going to be fleshed out at some point?

Automatic communication of trusted information for knowledge sharing.
Problems include a lack a standard protocol for information sharing, although there is some language development by programmers and researchers.

The reason i did this rearranging was because there was an incorrectly spelled link to "Artifical Intelligence" that contained most of the stuff that is now on the AI page. -- Amw

I'm not sure but I think I did most of that of that. The field was hyped earlier on but hasn't got too far. I want to know more about this topic and try to link it with topics like trust, knowledge, etc. I didn't get too far. I might of done some reading and written up some ideas or something. - ABliss
Defining AI as aiming to produce accurate computational models of human cognition is rather wide of the mark. I certainly would not want to produce an accurate model of human memory for example. I'd much prefer one that worked all the time. AI work in computer chess aims to perform better than humans and uses a very different approach to that of human cognition in order to achieve this. If you look at a typical AI conference proceedings, an AI text book or a university course in AI, you will find very little that supports the idea that AI is interested in "accurate computational models of human cognition". That is surely the goal of cognitive psychology. - GL7
Based on my limited expertise, I'd say that's accurate. While there are a variety of attempts to introduce organic computing into computers, the way we think is not something most are trying to reproduce, merely the ability to dynamicly adapt to surroundings and interact with sensors more effectively. All the research I read goes along the lines of: "we don't quite have the walking around bit complete yet." Even the army's unmanned drones crash on a fairly regular basis. However, this problem will unquestionably be solved in the next few years, and as computer systems grow in size and speed, I look forward to what their gradually increasing complexity develops. Maybe this will eventually become abstract thought or perhaps something unique. Webfork
Is your "we don't quite have the walking around bit complete yet." meant to be metaphorical  ?? I don't know what you mean. There are many military, scientific, medical and commercial AI applications in routine use today, and some have been around for years. - GL7
Yes, exactly. They can walk around and "do stuff" but they just aren't there yet. I think Sony came out with a robot that could climb stairs recently and, maybe its just me, but there is a bar in my mind that robots have not yet reached in terms of intelligence and reliability. i.e. the crash statistics on UMVs in the Air Force. Webfork