試験

問題 23 / 65

We are considering a pilot deployment in which AI automatically classifies emergency calls for emergency response, but using a model whose decision rationale is difficult to explain would be problematic. What are the main reasons such a model is unsuitable for introduction?

Cost reductions cannot be expected.
Safety and explainability cannot be guaranteed.
Additional servers are required.
Only the backlash from citizens is a concern.
It is mistakenly believed that the model is always accurate.

当サイトでは、ユーザー体験の向上を目的としてCookieを使用しています。サイトの利用を継続することで、Cookieの使用に同意したものとみなされます。