95% Generative AI Pilots at companies failing

Seems like from this article it points to a few reasons why AI is failing at companies:

Core issue being: Not the quality of the AI models, but the “learning gap” for both tools and organizations.
Companies find better success using pre-made AI solutions vs in-house made ones

I've found it to be good at summarizing things for me, but I find I lose the benefit of truly learning things when I use it.