Is AM going to die ?

Managing people's assets/money isn't just about the portfolio, that's quite literally like 1/4th of the true business model -- it's managing relationships, understanding their needs, knowing their tax situation, evolving to a changing world/market/economy. These are all things AI simply can't fully automate.

I'd go as far so to say that even traditional wealth management for wealthy families and UHNW individuals is never going to be fully automated, each investor's situations and goals are extremely unique. I still know plenty of regular middle-class individuals who are far more comfortable grabbing lunch with a local wealth manager and discussing how to allocate $50,000 for their kid's college fund or whatever; they're not that technologically advanced to trust these types of things on an algorithm. Why would this concept be any different on the institutional side? In fact, it's much more magnified. 

Let's just say we're focusing solely on the investment aspect of it, I highly doubt a pension fund or institutional investor is seriously going to trust $100s of millions of dollars on fully automated model -- go read into Knight Capital (Case Study 4: The $440 Million Software Error at Knight Capital - Henrico Dolfing)

I think anyone who tells you the entire investment management business is under threat to automation is either still a college student who hasn't truly ingrained themselves in the industry or is literally just flat out ignorant about how the business really works. 

Then there's the "passive investing always beats active investing" people -- these people slept through portfolio/investment analysis 101 in college, it's extremely inefficient to dump your entire investment portfolio into an index fund especially if you're UHNW, the market isn't just the SP500.