Alan, Your message had some worthwhile stuff in it, but it was rather long and covered many diverse topics. I'd rather your future AGI-list messages were more clearly focused.
I'll just respond to one point: You say: > Lets think of software as a pyramid. The simplest softwarez are at the > bottom, and AI is the little point at the very top... > > ^ < AI. > /_\ < applications. > /___\ < operating systems. > /_____\ < languages > /_______\ < raw computation. > > The ease of developing AI is a function of how well the lower levels > support the next level. This isn't strictly the case but given my state > of tiredness, it makes perfect sense. ;) > It is apparent to me that AI is, and has been for some time, strictly a > software problem. The time it will take to solve this software problem > will depend on how much overhead and complexity the AI researcher has to > deal wtith _BEFORE_ beginning his work. > > WE CAN MOVE THE SINGULARITY FORWARD BY YEARS JUST BY FIXING THE PROBLEMS > WITH TODAY'S SOFTWARE. There is a long history in AI of AI researchers turning into programming languages researchers. Because, "if we just had the right programming language, implementing AI would be really easy" !! Many of you will have read "Godel, Escher, Bach", where Doug Hofstadter made this claim very loudly and clearly: "AI Advances are Language Advances", he said. I've forced myself to resist this urge, because I've seen it lead a lot of people to spend their whole careers on programming languages research instead of AI research. I think that current computing infrastructure is awkward and conceptually misleading, for AGI research. However, I think that if you have a clear picture of the design you want to implement, the current computing infrastructure is adequate for implementing it. The problem is that AI experimentation is time-intensive using current computer software and hardware. Prototyping new ideas is time-consuming because the tools aren't "AI-natural." This does slow down development a lot, no doubt there. In short, better tools would be great to have, but their creation often proves a huge distraction from real AI work... -- Ben p.s. About your call for millions of dollars for your own research. Actually, many of us would like millions of dollars for our own research ;-> . My strong opinion is that posting one's need on e-mail lists is a VERY poor strategy for achieving such funding. ------- To unsubscribe, change your address, or temporarily deactivate your subscription, please go to http://v2.listbox.com/member/?[EMAIL PROTECTED]
