Premium
Reply to An inquiry into computer understanding
Author(s) -
Rendell Larry
Publication year - 1988
Publication title -
computational intelligence
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.353
H-Index - 52
eISSN - 1467-8640
pISSN - 0824-7935
DOI - 10.1111/j.1467-8640.1988.tb00108.x
Subject(s) - citation , mathematics , artificial intelligence , computer science , library science , information retrieval
are independent events, i.e., “shoot” being incapable of clipping “alive. ” Logicist: When you come down to it, the reason you ruled out “shoot” as a potential interference with “loaded” is because the two interact only if the victim was seen alive at r, and one of your independence laws says that interactions mediated via unconfirmed future events can be discounted. Isn’t this equivalent to Shoham’s scheme of “chronological ignorance” (Shoham 1986) whereby one sweeps forward in time and minimizes the number of abnormal events while ignoring, as much as possible, the effect of future events. Probabilist: Yes. The right to ignore unconfirmed future events is definitely a common feature of both schemes, but I am not sure at this point whether “chronological ignorance” captures all the context transformations licensed by the probabilistic interpretation of causality; the latter also teaches us how to manage facts that can’t be ignored by chronological considerations. Nevertheless, the logic of probabilistic independence does give Shoham’s scheme its operational and probabilistic legitimacy.