Through serendipitous web surfing I actually have a copy of their essay in Nature (now behind a paywall) from a few years ago that said that “The policy predictions of the models that are in use aren’t wrong, they are simply non-existant” As they reiterate in the video, “The models aren’t even good enough to be wrong”. With all the computer power available that can model so many inputs and variables it surprises me that economic models did not already attempt to make simulations of the economy in this way.