I think a lot of the time it’s about jumping on the hype train. “We’re going to use shiny new AI Agent for X” is easy to understand for higher-ups as progress. This feels pretty normal for progress.

I do think some of the underlying technologies have merit in certain areas, but imo they’re mostly in customer service/success where actual agents are overburdened and thus can’t provide a good customer experience. Knowledge regurgitation is something current tech (LLMs) are good at, so to me this makes sense. Feels like in the longer term more technical functions will start to work better once reasoning becomes better.