I had a professor that used to say that engineering is modelling problems with a set of differential equations and finding a numerical approximation to the solution. No differential equation ==> Not engineering.
The same expression "tocar ferro" was used traditionally in Catalan instead of knock on wood. Ferro translates as iron. Iron in this expression refers to a sword. More than trusting in good luck or some sort of supra natural protection, the expression "tocar ferro" conveys a sense of self reliance, of being prepared and confident.
I cannot read the paper that this article is based on, but it seems that it refers to the use of big data analytics and AI in 2024, not LLM. It concludes that the use of AI leads to a 4% increase in productivity. Nowadays the debate over AI productivity centers around LLMs, not big data analytics. This article does not seem to contradict more recent findings that LLM do not (yet) provide any increased productivity at the company level.
Grokipedia seems like a useful product from xAI to me. It is the first major change in how to write an encyclopedia since Wikipedia. On the topics that I have compared, Grokipedia entries are more comprehensive and relevant than Wikipedia ones. The only minus is that it lacks pictures. As an example compare the entries on the Norwegian lake Tinnsjå:
https://grokipedia.com/page/Tinnsj%C3%A5https://en.wikipedia.org/wiki/Tinnsj%C3%A5