I've worked in companies very similar to JPMorgan, and I've learned to be very skeptical about claims like this, because I've heard many.

Frequently, work hasn't been destroyed, only shuffled. Between departments (see, our forecasting workload has plummeted! In unrelated news, our model validation teams have grown!) or offshore (our worker-hours on this project are down! In unrelated news, we've initiated a contract with an offshore company that specializes in this type of work)

So, yeah, sure, maybe. I trust there is some interpretation of reality in which this investor-bait is technically true.

This makes a good niche use case for AI, estimating future cashflow is trying to consume a huge amount of regularly changing data to create forecasting. The vast majority of that job would be helped by a tool of this sort, but it still requires human input that knows the companies market and other contexts relevant to their company and specific industry.

It sounds like its a big deal but its really not. Even in billion dollar companies there's usually only one person doing this job.

Alternative take: I'm glad to see them acknowledge how little value there is to their work. Maybe soon we'll be able to replace the whole of JPMorgan with an AI chatbot then.