Computational Difficulties

Attempts to build systems based on probability theory were started in the 1960’s.
This research was abandoned because:
- calculation load was considered intractable Gorry (1973)
- probabilities were thought to be ‘epistemologically inadequate’ McCarthy and Hayes (1969)
- it was claimed that
- ‘people are bad probability estimators’
- ‘use of probability requires massive amounts of data’