mclarke@ac.dal.ca (10/22/90)
After having posted the following, I see a need to clarify my point. Time and progress march on ... Technology advances and society is adjusting to the new technologies and incorporating it. ...We take gradual change in our lives for granted and tend to accept incremental changes without question. If this is not handled now there will be little time to react when it takes effect. There are two evolving technologies which will gradually change the way society makes day to day choices. They are artificial intelligence and parallel processing. Essentailly this means that within about 20 years computer technology should be able to emulate the human reasoning process.. ...Texas Instruments, among others claim that "quantum chips" are soon to be a reality and expect production near the year 2000. (Scientific American, Quantum Chips). Since there is about a ten year lag time between introduction and full application of a technology that plants us around the year 2010 for implementation. The massive change in size of chips and potential sophistication suggests that parallel processors may be available near that time which approach the human mind in reasoning for restricted applications. ...Artificial Intelligence software (Expert Systems) which attempts to emulate the human reasoning process is now restricted by the current power of computer equipment. Even Cray supercomputers are becomming restrictive in large AI or Expert applications. When the new chip technology becomes available a Cray could fit in a desktop computer. It is not inconcievable that companies like Cray, IBM, Etc. will have extremely powerful applications of quantum chips in new supercomputers which can provide a suitable platform for AI applications. When the technologies mature, corporate and government choices can be made by computers. They will be more efficient, faster, more accurate and much cheaper. Professional jobs will be done by computer. This includes medical, legal, etc. positions. ...Many qualitative choices wiil be made by computer. For example, medical diagnosis (already a reality), traffic court judgements. The question is " WHO is going to teach the computers human values, morals, etc." Christian bodies had better start investing $$$ in research now to avoid future crisis. ...You can teach a human being values but who will be responsible for creating a set of values for computer judgements? Should the Church not become an advocate for creating a worldwide "standard"? ...I am not looking for "soft dollars". I don't want to see a crisis in society that could have been prevented by persons more "expert" (no pun intended) than I. Michael Clarke Halifax, N.S. Des Colores
kwilson@urbana.mcd.mot.com (Kent Wilson) (10/23/90)
In article <Oct.22.02.05.27.1990.19950@athos.rutgers.edu>, mclarke@ac.dal.ca writes: |> |>The question is " WHO is going to teach the computers human values, morals, |>etc." Christian bodies had better start investing $$$ in research now to |>avoid future crisis. |> |>...You can teach a human being values but who will be responsible for |>creating a set of values for computer judgements? Should the Church not |>become an advocate for creating a worldwide "standard"? |> Oh c'mon!!! Haven't you heard of Separation of Church and AI Software? :) Go read your Constitution! |> |>Michael Clarke |>Halifax, N.S. |> |>Des Colores Kent ===============================================================================
mmh@cs.qmw.ac.uk (Matthew Huntbach) (10/25/90)
In article <Oct.22.02.05.27.1990.19950@athos.rutgers.edu> mclarke@ac.dal.ca writes: >There are two evolving technologies which will gradually change the >way society makes day to day choices. They are artificial intelligence >and parallel processing. Essentailly this means that within about 20 years >computer technology should be able to emulate the human reasoning process.. > I happen to be a researcher in the fields of "artificial intelligence" (I don't like the term myself - I prefer something like "experimental computing") and parallel processing. From my experience I believe this is absolute nonsense. It is not possible to see where anything remotely human-like could even come from. >computers. They will be more efficient, faster, more accurate and much >cheaper. Professional jobs will be done by computer. This includes >medical, legal, etc. positions. A computer simply obeys rules. That's it. What you are really complaining about is a rigid adherence to rules. There is no need to bring computers into it - exactly the same situation would arise if doctors and lawyers always stuck rigidly to a fixed set of rules. Doctors and lawyers don't usually behave in this way though, they bring in their own human knowledge to interpret the rules, and deal with special cases that weren't considered when the rules were drawn up. In this way they can come up with better solutions. However humans can go against the rules by mistake, which is where use of computers has an advantage. Rather than confuse the issue by bringing in "artificial intelligence, you should be thinking in more practical terms. Matthew Huntbach
kwb@hpmtlx.lvld.hp.com ($Keith_Blackwell) (11/05/90)
/ hpmtlx:soc.religion.christian / mclarke@ac.dal.ca / 12:05 am Oct 22, 1990 / | |...Artificial Intelligence software (Expert Systems) which attempts to emulate |the human reasoning process is now restricted by the current power of |computer equipment. Having done some research into cognitive modelling while in Graduate School, I believe this statement is a gross misunderstanding. Sure, the "current power of computer equipment" may be a limiting factor in some sense, and that is one factor that is quickly changing. But the most important limiting factor is our own ignorance of the complexities of the human cognitive processes. That factor is not going to change rapidly. Some fascinating theories have come out and been refined in the past 30 years, and many of those could be combined and implemented in some emulation system running on super-hardware alluded to later in the above post. But the very practical problem of how to get such a system rolling would take enormous person-resources to solve. Having gotten a vision for the kind of research necessary to accomplish the goal of having person-emulation similar to what we find in science fiction, I am convinced that we won't see any actual working models come anywhere close for another 70 or so years. Technology can increase a million fold, and we still don't know what do with that technology. The questions we need answers to simply cannot be answered within the next 20 years. If you are interested in this field, I suggest you browse through issues of the _Cognitive_Science_ journal at a research library. As for morality, others have answered that issue well (and not so well). This may become an issue when (connectionist?) systems begin to take on responsibilities within the realm of interpersonal relationships, but I don't think that will happen for many decades. --- Keith Blackwell