[Limits to Science, Steven Rose]
p.31-33 The point is that the mode of thinking which has characterized the period of the rise of
science from the seventeenth-century is a reductionist one. That is, it believes not merely that to understand the world requires
disassembling it into its component parts, but that these parts are in some way more fundamental than the wholes they compose...
The fallacies of such reductionism should be apparent... Yet reductionism runs deep... So long as science - in the
questions it asks, and the answers it accepts - is couched in reductionist and determinist terms, understanding of complex
phenomena is frustrated... Failing to approach the complexity of such systems, reductionism resorts to more
or less vulgar simplifications
[Artefactual Intelligence, Patrick Wall and Joan Safran]
p.120 A computer revolution is in progress which is changing the world. It depends on machine
ability to handle precisely defined tasks with enormous speed, capacity and reliability.
p.121 The intelligent animal not only detects order in a disordered world but detects order which
has a biological meaning in terms of survival. Furthermore it devises alternate strategies to move from the observed present
to the desired future.
p.128 Two related aspects separate the intelligent perception of all mice and humans from the performance
of machines.
(1) Intelligent perception always takes place in the context of both immediate and historical
perspective, and involves selective attention.
(2) Intelligent perception always takes place in the context of goals and actions and involves meaning
expressed in terms of the biological needs of the whole organism.
p.128-129 selective attention is a dominating aspect of our mental activity and living
creatures show clear signs of the selectivity. Our sense organs are continuously bombarded with a barrage of data about the
details of the world around us. Somehow we continuously reject 99.9% as irrelevant and select out those features which
are of importance to our problems of the moment. If there is one mark of intelligence it is this selective ability...
Of course, with hindsight we can sometimes tell what are the relevant clues and then there is no problem in teaching
them to a machine or a student.
[Minds, Machines and Meaning, Richard Gregory]
p.132 computers... cannot understand symbols (or indeed
anything else either), though they can manipulate symbols, according to formal rules, with consummate speed and accuracy far
surpassing our own fumbling efforts... they do not understand the questions they are asked or the answers they provide.
At least for the present... they do not understand.
p.133 Personally I would accept a machine as intelligent if it can cope appropriately with novelty.
