An informational learning process fills existing buckets with more knowledge. Say math, computer science or sociology.
When learning transformationally we create new buckets. It is not so much about the learning of the material itself, but the learning that the material exists. By learning of the existence of these buckets we create mental models about new connections we didn’t have before.
Machines are of course infinitely capable of the first, and notoriously useless in the latter.
p.s. we can also think of informational learning as answers, and transformational learning as questions