An Issue with Artificial Intelligence
Sunday, July 22, 2018 at 02:35PM
Charlie in Becoming an XBRL Master Craftsman, Digital Financial Reporting

Artificial intelligence has an issue.  You need to be aware of this issue because it means that artificial intelligence should not be employed for solving certain classes of problems.  The issue is straight forward.

There are two types of reasoning: deduction and induction. These two videos help you understand the difference between deduction and induction: Video 1; Video 2.

The bottom line is that deduction is certain; you can be logically certain of the conclusions reached using deduction. That is the goal of deduction, certainty. Induction is not certain, induction is based on probability.

The Black Swan Problem of Induction clearly explains the problem of induction.

“No amount of observations of white swans,” he said, “can allow the inference that all swans are white, but the observation of a single black swan is sufficient to refute that conclusion.”

Many employing artificial intelligence use inductive reasoning systems.  They believe that these systems can work because the problems caused by induction can be managed effectively.  That may be true.  But it likewise may not be true for certain knowledge domains.  The call here should be made by domain experts, not computer scientists that do not have a deep understanding of the domain.

#############################

One problem to explain why AI works

Article originally appeared on Intelligent XBRL-based structured digital financial reporting using US GAAP and IFRS (http://xbrl.squarespace.com/).
See website for complete article licensing information.