The error of futurism: prediction and computer ethics

Horner, David (2003) The error of futurism: prediction and computer ethics ACM SIGCAS Computers and Society, 33 (4). p. 1. ISSN 0095-2737

Full text not available from this repository.


The aim of this paper is to challenge the assumption that we can insure ourselves against natural and moral evils through improvement in our powers of prediction. I argue that computer ethicists need to take account of the epistemological and logical limits to our knowledge of the future. The supposition that we can accurately predict future events is 'the error of futurism'. At least part of the job of Computer Ethics has been seen to be the need to foresee and forestall the undesirable and unintended consequences of computer technologies. The paper seeks to show that our faith in the power of prediction is ill-founded and is often based on a misconceived picture of the future. There are reasons for supposing that the limits on prediction are not only epistemological but also logical. The epistemological problems are illustrated by reference to: 'information effects' (problems with information gathering and processing); Oedipus effects (the effect of making a predictive statement about the human situation to which that statement refers); and 'revenge effects' (unintended, undesirable, unpredictable consequences of innovation). The logical problem with prediction is that when we refer to 'future events' we don't refer to events at all. The phrase 'future event' is not a descriptive phrase that characterises some special kind of event as the phrases 'sudden event' or 'historic event' do. 'Future event' refers to an unrealised possibility until it comes to pass. The Y2K phenomenon is illustrative of some of these problems of accurate prediction. We can use our knowledge of the past (and we surely can talk about 'our knowledge of the past') as a guide to action. This does not imply that the future will necessarily be like the past only that it is reasonable to proceed on this basis and that prediction must play a much more limited role in our thinking. It can hardly be moral to claim to know what we cannot know.

Item Type: Journal article
Subjects: G000 Computing and Mathematical Sciences
DOI (a stable link to the resource): 10.1145/968358.968359
Faculties: Faculty of Science and Engineering > School of Computing, Engineering and Mathematics > Social Informatics
Depositing User: editor cmis
Date Deposited: 04 Dec 2006
Last Modified: 24 Mar 2015 12:31

Actions (login required)

View Item View Item


Downloads per month over past year