gilbert@hci.hw.ac.uk (Gilbert Cockton) (01/29/88)
In article <4222@utai.UUCP> tjhorton@ai.UUCP (Timothy J. Horton) writes: >>...People who flounder hopelessly are probably short on their >>philosophical training. > >Not true. Realize, also, that there are conceptual chasms between fields. > >Philosophical arguments about computational models of intelligence, for >instance, among those without comprehensive conceptual bases in computer >science, often seem to reduce to expressions of superstition and ignorance, >at least among the vocal. On conceptual chasms, what - philosophical analysis apart - can bridge them? On ignorance of computability in 'philosophical' arguments on natural and artificial intelligence, perhaps the Theory of Computation needs to be as much a part of a proper philosophical training as linguistic analysis and formal logic. Some people in AI could do with it as well (i.e. those who don't have it). As for reduction to superstition, isn't this the outcome for an analysis of many 'natural truths'. On the existence of objects, nothing is 'proven', but nevertheless, we find no reason for rejecting the natural truth of their existence. Arguments based on ignorance must be discounted, but are we not left with the case that we still have no reason for rejecting the natural truth that human and machine intelligence are different? Not only is the case for the equivalence of human and machine intelligence not proven, no analysis exists, to my knowledge, which points to a way of establishing the equivalence. This leaves AI as a piece of very expensive speculation based on beliefs which insult our higher views of ourselves. Superstition no doubt, but a dominant and moral superstition which needs to command some respect. Vocal polemic is as much a reaction to the arrogant unreasonableness of some major AI pundits, as it is a reflection of the incompetence of the advocate. The debate has been fair on neither side, and the ability of AI pundits to stand their ground is due to their social marginality as round-the-clock scientists and their cultural marginality as workers outwith a proper discipline (look up Sociology of Deviance). People who live in bunkers don't get hit by stones ;-) The big AI pundits just remind me of Skinner. BTW: NOT(AI pundit = AI worker) - most AI workers know their systems aren't working (yet?) and do leave their bunkers to mingle :-) >I suggest, in balance, Russell's "The Cult of Common Usage," for instance. Great - keep balancing, more competent philosophy for the reading list. >Experience would seem to indicate that a few vocal individuals may press >their arguments on the entire network, rather than delivering ambivalent >analysis or investigating before disseminating. Sounds like a netiquette proposal which I thoroughly endorse. Whilst guilty of advocacy on occasions, I think that everyone should strive for an ambivalent analysis in this sort of public forum, and leave people to make their own minds up. Sounds like good philosophy to me. However, ambivalence cannot be expected in response to imcompetence, however candid. Witness the current debate on economic structure and diachronic syntax. Nor, as with tolerance of the intolerant, I can't be ambivalent about dogmatists. -- Gilbert Cockton, Scottish HCI Centre, Heriot-Watt University, Chambers St., Edinburgh, EH1 1HX. JANET: gilbert@uk.ac.hw.hci ARPA: gilbert%hci.hw.ac.uk@cs.ucl.ac.uk UUCP: ..{backbone}!mcvax!ukc!hci!gilbert