AI assistants are increasingly embedded in daily workflows:
coding, writing, reasoning, decision support.
They clearly reduce local cognitive effort.
But an open question remains:
do they also accumulate a form of delayed cognitive cost?
This work proposes a formal, non-narrative framework to model
the trade-off between immediate effort reduction and long-term
dependency effects, referred to here as “cognitive debt”.
The focus is strictly structural:
decision transfer, reversibility, control retention,
and long-term cognitive autonomy.
No ethics essay. No hype.
A model meant to be inspected, challenged, or discarded.
Full framework:
https://osf.io/ub5f4
Top comments (0)