Mike, On 5/12/08, Mike Tintner <[EMAIL PROTECTED]> wrote: > > Reading the following in a psych. group, made me think of your diagnostic > program - the real world is generally messy: The principle applies to all > AGI programs. > > "there is of course an interesting phenomenon in medicine generally where > theories of causality may swing with the development of new data > > What I find most interesting is "negative causality", where a flawed study dismisses something that is VERY important, and in the process stymied entire fields for decades.
We see this in AI/AGI with Joe Weizenbaum's dismissal of his own shallow parsing methods, which work fantastically well for many things, provided of course that shallow parsing is NOT combined with Eliza-like word manipulation. Joe threw the baby out with the bathwater, and AI still hasn't fully recovered. The same thing has happened in several areas of medicine so that, for example, ~100% of all cancer research is a COMPLETE waste of effort. - consider things like stomach ulcers, generally seen as a purely medical > problem > > ... in a treatment-only paradigm. Until recently, no one ever talked about curing them. until theories of stress (and especially Brady's work) had us all > believing this was the key. > > And well it may have been, as stress is immunosuppressive, potentially allowing helicobacter (or other yet-to-be-discovered pathogens) to grow. Then we had helicobacter come onto the scene, and everything became > medical again > > But, at least we could cure some cases. - until we discovered that we still had to explain why many people had > helicobacter but no sign of ulcers..... > > Nonetheless, it sure doesn't hurt to kill the helicobacter since this DOES cure a lot of cases. Note however that the people whom this does help may also have suppressed immune systems, which is a separate problem that should also be corrected. Story: On a routine blood test, I once noticed a low ANL=1500 (low neutrophil count, the most common type of white blood cell), a stress indicator that was low enough that, for example, I could not receive any medications that might potentially lower it further, though I had no direct symptoms of a low ANL. A week of research, talking to experts, etc., and I had a workable theory, that my high espresso consumption was the problem. I cut my espresso consumption in half, waited a week and was retested, and my ANL doubled back to a more healthy value. More often than not, illnesses are complex with intersecting cause-and-effect chains, with the major problems being at the intersections. Is often *far* from easy to identify causes or even to specify "one" > cause." > > Because of course, there is seldom a single cause. Further, causes may be now long gone, with only the self-sustaining loops at the ends of cause and effect chains causing all of the problems. - - - Of course, the BIG (in an AGI sense) thing is to do the best job possible with as little information as possible. In the real world, most diagnoses (including stomach ulcers) are for groups of a dozen or so illnesses that all have very nearly the same major symptoms. What might cure one particular illness may actually make another worse, for no net statistical effect in the pseudoscientific "double blind" studies. "Modern" "science" has just now progressed to the point that it is possible to reverse engineer most illnesses and discover their primary cause and effect chain(s). Doctors don't do this because it is still a LOT (days if you are lucky, and months if you are not so lucky) of work and they lack the skills to succeed, but I have personally done this for a number of people's illnesses. Once you have done several illnesses and characterized each of the links that are involved, these links will also be on some entirely different chains for apparently quite different illnesses, so the job gets a LOT easier after doing a few illnesses. I estimate that it would only take ~100 man-years to do this job for all major illnesses, thereby providing true life-long cures for most of them. This would require but a minuscule fraction of the present medical research budget, but apparently this will NEVER happen because it would doom the drug companies that now so completely control all current medical research. If you know anyone with a lot of money and who wants to do some REAL research, then please send them my way. Steve Richfield ------------------------------------------- agi Archives: http://www.listbox.com/member/archive/303/=now RSS Feed: http://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: http://www.listbox.com/member/?member_id=8660244&id_secret=101455710-f059c4 Powered by Listbox: http://www.listbox.com
