Minds and Computers : An Introduction to the Philosophy of Artificial Intelligence
Download 1.05 Mb. Pdf ko'rish
|
document (2)
algorithms. We will call an algorithm a learning algorithm if, as a result
of employing it, the system is conferred with greater capacities. There are at least two ways in which a mind can be conferred with greater capacities – it can gain new information or it can attain, or improve, skills or abilities. In the former case, the mind is learning that things are the case. In the latter, the mind is learning how things are done. In the system [MIND] learning that things are the case corresponds with storing new content – i.e. storing new values in registers (where these values code content). Learning how things are done corresponds with storing new algorithms or with optimising existing algorithms. Now, given our understanding of [MIND] as a kind of [OS] which is in continual iterated operation and which is expandable – i.e. has algo- rithms which store content and which generate or optimise further algorithms, both of which confer on the system greater capacities – we 105 can help ourselves to an explanation available to the computationalist of variation among [MIND]s with respect to capacities. Given that, as already established, any two [MIND]s are highly likely to be in distinct states, and given that, depending on the amount of time it has been operating for and the nature of the inputs it has received, a [MIND] will contain more or less stored content and greater or fewer available stored algorithms (optimised better or worse), it should come as no surprise that any two [MIND]s will vary significantly with respect to capacities. This explanation responds to one of the two challenges presented to the computationalist by the learning objection – it explains how two isomorphic formal systems can have di fferent capacities and, hence, how minds can be held to be isomorphisms of [MIND] despite immense variation among minds with respect to what they can do. The challenge remains, however, for the computationalist to specify algorithms which govern learning. This is a challenge to which we will return at various points in the following chapters, particularly in Chapter 13 when we discuss automated reasoning systems, and in Chapter 19 when we examine learning in artificial neural networks. We will also be investigating the way in which humans learn languages in Chapter 16 and considering evidence that this learning is rule governed. We have now considered two objections one might mount against computationalism. These were essentially stronger and weaker ver- sions of the same objection – minds vary. In both cases, we have seen how a computationalist might reasonably respond. Let’s consider one further objection against the theory that minds are computational devices. 10.5 CREATIVITY Another standard prima facie objection appeals to the human creative capacity, as follows. The operations of formal systems are entirely mechanical but minds are creative. Minds create great works of art, music, architecture and literature, and have an enormous capacity to innovate. This characteristic creativity of human minds seems to be compelling evidence against computationalism which seeks to account for mentality in terms of purely mechanical operations. It is certainly the case that it seems that nothing could be further from an algorithmic process than painting an artwork or composing an orchestral symphony. As we saw in Chapter 2 however, the way things seem is no reliable indicator of the way things are. 106 The challenge here for the computationalist is to explain how the mental functions we cite as paradigmatically ‘creative’ can be algo- rithmically delivered, contra-intuition. For that explanation, we need an understanding of this notion of creativity. The opponent of computationalism might endorse a definition of creativity along the lines of: an activity is creative if its result is the pro- duction of a work (an artwork, composition, etc.) which could not have been produced by simply following a rule-governed procedure. Although this definition is somewhat intuitive – after all we’re all fairly certain da Vinci wasn’t painting by numbers when he painted the Mona Lisa – it begs the question rather straightforwardly against the computationalist. Whether or not creativity can be accounted for algorithmically is precisely what is at issue. So what is it about creativity such that our initial intuition is to con- trast creative behaviour with rule-governed behaviour? Well, firstly, not everyone is equally creative. People have di fferent capacities for engaging in creative enterprises. We have already seen, though, how a computationalist can account for variation with respect to capacities so this is not su fficient as an objection, but it points us in the right direction. It seems that what disposes us initially against the notion of cre- ative behaviour being rule-governed is an intuition that were it rule- governed, it would be more readily teachable. It is characteristic of those we laud as creative masters – artists, artisans, composers, etc. – that there is something mysterious to others about their talent. Further, it seems that when it comes to creative endeavour, one either ‘has it’ or not. Certainly one can learn various techniques and methods for working with materials to generate certain e ffects; however, it is not clear how one could learn to ‘be creative’ per se. This does not, though, speak against the possibility that this kind of behaviour is indeed underwritten by computational processes. Certainly the opponent of computationalism may well demand an account of how such behaviour could be computationally delivered; however, there are certain responses available. For instance, the computationalist might tell a story about certain algorithms requiring certain computational resources for their imple- mentation, such that variation with respect to the ability to acquire certain algorithms is to be explained in terms of variation in the sub- strate (brains) in which [MIND]s are realised. In any case, the computationalist is well within their rights to dissent from answering the question – how is creative behaviour com- putationally delivered? – until this notion of ‘creative behaviour’ is 107 more rigorously specified. In the absence of an independently plausi- ble account of creativity which speaks directly against the possibility of such behaviour being rule-governed, the computationalist is no worse o ff in this respect than other theorists. We have now considered a number of preliminary objections against computationalism and have seen how, in each case, the objection fails. We have yet to consider more sophisticated arguments against the theory. We shall hold o ff on these until after we have seen some arti- ficial intelligence applications which will provide context for phil- osophical objections. To recapitulate, we have discussed some common misunderstand- ings of computationalism and some bad arguments against the theory which trade on these misunderstandings. We have advanced a Download 1.05 Mb. Do'stlaringiz bilan baham: |
Ma'lumotlar bazasi mualliflik huquqi bilan himoyalangan ©fayllar.org 2024
ma'muriyatiga murojaat qiling
ma'muriyatiga murojaat qiling