[ativzzz]

That depends on the definition of

> psychological harm

Are they going to go into a codebase and figure this out? If I use AI to pull these levers vs using a series of if/else statements:

  if person.has_bought_multiple_items

    increase_base_price(person)

that's legal I guess?

[p_l]

The wording looks to me like it's designed to combat the feedback loop of "observe user behaviour" -> "apply ML model" -> "reoptimize addiction loop"