This is definitely an "old man yells at clouds" post. The author seems to believe that future AI-driven investment systems will work like ChatGPT and so will need to be "talked" to, which is bizarre. By 2040, human financial planners will have gone the way of the human computers of the 1940s.

Me neither. Besides, there is this enduring confusion all over the place that "AI := LLM". I'd expect the radical AI (ML) innovations to be somewhere else in the long-run (like, say, in robotics, as hinted by the title but downplayed by the text).