> The truth is that the development of complete AI systems is extremely challenging and the organizational effort required to drive such projects to completion is orders of magnitude higher than developing proof-of-concept or prototype models.

This is very true, and I agree very much with Michael Jordan's thoughts (referenced in the article) that AI has yet to become a serious engineering discipline, and that only through this happening will we see substantial impacts of AI in meat-space.

At the moment this is feasible for big tech companies (see e.g. autonomous driving) but I'm interested to see what might happen when tools are developed that make this possible for nimbler organisations