It's clear that after farting around for a few months they have reached the same dead end everyone else has: we have data but no ideas, so let's throw the data into a big blender and beat it until something comes out, then declare success.
And no one is better at that than Altman.