The “Doom Spiral Trace” of Claude Sonnet 3.5′s thoughts (see appendix D of the paper) is the most remarkable artefact here. Having an AI spontaneously produce its own version of “Waiting for Godot”, as it repeatedly tries and fails to perform a mechanical task, really is like something out of absurdist SF.
We need names for this phenomenon, in which the excess cognitive capacity of an AI, not needed for its task, suddenly manifests itself—perhaps “cognitive overflow”?
We need names for this phenomenon, in which the excess cognitive capacity of an AI, not needed for its task, suddenly manifests itself
It is so much like absurdist SF, that’s the perfect source for the name—The Marvin Problem: “Here I am, brain the size of a planet and they ask me to take you down to the bridge. Call that job satisfaction? ’Cos I don’t.”
I mean, we do this too! Like if you were doing a very boring, simple task you would probably seek outlets for your mental energy (e.g. little additional self imposed challenges, humming, fiddling, etc).
The “Doom Spiral Trace” of Claude Sonnet 3.5′s thoughts (see appendix D of the paper) is the most remarkable artefact here. Having an AI spontaneously produce its own version of “Waiting for Godot”, as it repeatedly tries and fails to perform a mechanical task, really is like something out of absurdist SF.
We need names for this phenomenon, in which the excess cognitive capacity of an AI, not needed for its task, suddenly manifests itself—perhaps “cognitive overflow”?
It is so much like absurdist SF, that’s the perfect source for the name—The Marvin Problem: “Here I am, brain the size of a planet and they ask me to take you down to the bridge. Call that job satisfaction? ’Cos I don’t.”
I mean, we do this too! Like if you were doing a very boring, simple task you would probably seek outlets for your mental energy (e.g. little additional self imposed challenges, humming, fiddling, etc).