Bundy%edxa@ucl-cs.arpa (06/20/84)
From: BUNDY HPS (on ERCC DEC-10) <Bundy%edxa@ucl-cs.arpa> Credibility has always been a precious asset for AI, but never more so than now. We are being given the chance to prove ourselves. If the range of AI products now coming onto the market are shown to provide genuine solutions to hard problems then we have a rosy future. A few such products have been produced, but our future could still be jeopardized by a few, well publised, failures. Genuine failures - where there was determined, but ultimately unsuccesful, effort to solve a problem - are regretable, but not fatal. Every technology has its limitations. What we have to worry about are charlatans and incompentents taking advantage of the current fashion and selling products which are overrated or useless. AI might then be sigmatized as a giant con-trick, and the current tide of enthusiasm would ebb as fast as it flowed. (Remember Machine Translation - it could still happen.) The academic field guards itself against charlatans and incompentents by the peer review of research papers, grants, PhDs, etc. There is no equivalent in the commercial AI field. Faced with this problem other fields set up professional associations and codes of practice. We need a similar set-up and we needed it yesterday. The 'blue chip' AI companies should get together now to found such an association. Membership should depend on a continuing high standard of AI product and in-house expertise. Members would be able to advertise their membership and customers would have some assurance of quality. Charlatans and incompetents would be excluded or ejected, so that the failure of their products would not be seen to reflect on the field as a whole. A mechanism needs to be devised to prevent a few companies annexing the association to themselves and excluding worthy competition. But this is not a big worry. Firstly, in the current state of the field AI companies have a lot to gain by encouraging quality in other companies. Every success increases the market for everyone, whereas failure decreases it. Until the size of the market has been established and the capacity of the companies risen to meet it, they have more to gain than to lose by mutual support. Secondly, excluded companies can always set up a rival association. This association needs a code of practice, which members would agree to adhere to and which would serve as a basis for refusing membership. What form should such a code take, i.e. what counts as malpractice in AI? I suspect malpractice may be a lot harder to define in AI than in insurance, or medicine, or travel agency. Due to the state of the art, AI products cannot be perfect. No-one expects 100% accurate diagnosis of all known diseases. On the other hand a program which only works for slight variations of the standard demo is clearly a con. Where is the threshold to be drawn and how can it be defined? What consitutes an extravagent claim? Any product which claims to: understand any natural language input, or to make programming redundant, or to allow the user to volunteer any information, sounds decidedly smelly to me. Where do we draw the line? I would welcome suggestions and comments. Alan Bundy
abc@brl-tgr.UUCP (06/27/84)
I suggest that the ACM provides an appropriate umbrella under which such an effort can at least be planned. It is sufficiently broad-based as to be representative and not exclusive and its democratic procedures provide protection from the types of abuses that could be possible. (I do not mean to slight the AAAI; it's just that ACM seems to have more of the "mechanisms" that such an efort will need.) Also, I have felt for many years that ACM should, at least in the US, provide the kind of accreditation of Computer Science curricula that the engineering societies provide for theirs.