Let’s talk about the Turing test! The primary objective of a text generator is to predict a sequence of words related to a given sentence. It won’t remember the context of a discussion, so it’s bound to fail the test.
Text consistency also remains a barrier to successful testing, depending on the data collected. A varied and dense corpus will give more consistency to an AI.
Press articles and literary works are protected by copyright. They must be acquired in order to be used, and their use may be limited.
Some works, however, have entered the public domain, and can therefore be used by anyone. But they are witnesses to their times, and can no longer correspond to contemporary values.
