Interesting read, thanks! Perhaps the word "meaning" should have been used more than once in the article, as it may help explain the axiological aspects of the problems discussed.

Your self-reference in AI reminded me of Brandom-ian approach to AI. Here is an article I found ( https://arxiv.org/pdf/1803.02912.pdf ), but I think it's worth reading the Brandom himself from the bibliography in the article.