, Volume 2, Issue 1, pp 3–16 | Cite as

Pathologies of AI: Responsible use of artificial intelligence in professional work

  • Ronald Stamper


Although the AI paradigm is useful for building knowledge-based systems for the applied natural sciences, there are dangers when it is extended into the domains of business, law and other social systems. It is misleading to treat knowledge as a commodity that can be separated from the context in which it is regularly used. Especially when it relates to social behaviour, knowledge should be treated as socially constructed, interpreted and maintained through its practical use in context. The meanings of terms in a knowledge-base are assumed to be references to an objective reality whereas they are instruments for expressing values and exercising power. Expert systems that are not perspicuous to the expert community will lose their meanings and cease to contain genuine knowledge, as they will be divorced from the social processes essential for the maintenance of both meaning and knowledge. Perspicuity is usually sacrificed when knowledge is represented in a formalism, with the result that the original problem is compounded with a second problem of penetrating the representation language. Formalisms that make business and legal problems easier to understand are one essential research goal, not only in the quest for intelligent machines to replace intelligent human beings, but also in the wiser quest for computers to support collaborative work and other forms of social problem solving.


AI paradigm knowledge semantics complexity responsibility collaboration law 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. Bench-Capen, T. (1987). Social Security Expert Systems in Göranzen (1987a).Google Scholar
  2. Cooley, M. (1987).Architect or Bee? The Human Price of Technology. The Hogarth Press, London.Google Scholar
  3. Dreyfus, H.L. and S.E. Dreyfus (1986).Mind Over Machine. Blackwell, Oxford.Google Scholar
  4. Ennals, R. (1987). Action Research Case Studies in Göranzon (1987a).Google Scholar
  5. Gill, K.S. (1987). Knowledge and Skill Transfer: New Technology and Issues of Training in Göranzon (1987a).Google Scholar
  6. Göranzon, B. (ed.) (1987a). Proceedings of the Symposium onCulture, Language and Artificial Intelligence. 2 vols. Swedish Centre for Working Life, Stockholm.Google Scholar
  7. Göranzon, B. (1987b). The Practice of the Use of Computers,AI & Society. 1,1. 25–36.Google Scholar
  8. Guller, P. (1987). Automation Skill and Apprenticeship in Göranzon (1987a).Google Scholar
  9. Hilton, J. (1987). Numinous Knowledge: Some Thoughts on Interactive Video and Expert Systems in Göranzon (1987a).Google Scholar
  10. Josephson, I. (1987). The Nurse as an Engineer,AI & Society. 1,2. 115–126.Google Scholar
  11. Negrotti, M. (1987). The Piping of Thought,AI & Society. 1,2. 85–91.Google Scholar
  12. Sergot, M.J., F. Sadri, R. Kowalski, F Kriwaczek, P. Hammond and H.T. Cory (1986). The British Nationality Act as a Logic Program,Communications ACM. 29,5. 370.Google Scholar
  13. Stamper, R.K. (1985). Management Epistemology: Garbage In, Garbage Out in Methlie and Sprague (eds)Knowledge Representation for Decision Support Systems. Amsterdam.Google Scholar
  14. Stamper, R.K. (1986). Fitting a Technology to a Society: Legol/Norma Paper L97. London School of Economics, London.Google Scholar
  15. Stamper, R.K., K. Althous and J. Backhouse (1987a). Legol/Norma Project: A Technical Survey: Legol/Norma Paper L98. London School of Economics, London.Google Scholar
  16. Stamper, R.K. (1987b). Semantics in Boland and Hirschheim (eds)Critical Issues in Information Systems Research. John Wiley and Sons, Chichester.Google Scholar

Copyright information

© Springer-Verlag London Limited 1988

Authors and Affiliations

  • Ronald Stamper
    • 1
  1. 1.Department of Systems AnalysisLondon School of Economics and Political ScienceLondonUK

Personalised recommendations