While not exactly parables, the AI Koans in HAKMEM (MIT AI Lab memos) / the Jargon File may be considered a subtype.
A Selection of AI Koans: The MIT jargon file version 299
See also, Hacker koan - Simple English Wikipedia, the free encyclopedia
E.g.,
In the days when Sussman was a novice, Minsky once came to him as he sat hacking at the PDP-6.
“What are you doing?”, asked Minsky.
“I am training a randomly wired neural net to play Tic-Tac-Toe” Sussman replied.
“Why is the net wired randomly?”, asked Minsky.
“I do not want it to have any preconceptions of how to play”, Sussman said.
Minsky then shut his eyes.
“Why do you close your eyes?”, Sussman asked his teacher.
“So that the room will be empty.”
At that moment, Sussman was enlightened.
Again, this is a bit of a double-layered affair, since Minsky isn’t just hinting Sussman at a random configuration still being a configuration, but this is also Minsky, who initiated the end of the first boom of neural networks by providing the proof for their intrinsic XOR problem (which was [somewhat; personal musing] overcome by multi-layer networks eventually). While probably not intended by its original reading, as Minsky thoughtfully shut his eyes, there was no network and no preconception, and the (conceptual) room was empty, indeed.
P.S.: Sussman’s TicTac-Toe for the PDP-6 would have been displayed on a Type 340 scope, compare this thread, LIFE - 4 Gosper Glider Guns on PDP-7 Type 340 display