So glad you asked. I have been following it closely since the early fall, and so far, it is the best model. Granted, my analysis is for our little area only, and is entirely empirical. Here is a little anecdotal example:
I am also active on a Notre Dame football board, where I have installed myself as the weather nerd in-chief. About 5 days ago, it appeared likely that we would be hosting Alabama on the 20th. Of course, I went to the maps hoping for the most miserable cold possible for northern Indiana in December. At that time, gfs and euro showed a nationwide zonal flow, with temps far above normal for SBN. Euro AI on the other hand, had deep winter. I am not ready to spike the ball, since we are only on the 15 yard line, but AI has held fairly steady while the others are showing signs of capitulation. We'll see what happens!
I am ready to make this prediction: Next year at this time we will all be going to the AI models first...with occasional references to the "legacy models".