David RevoyThe first dose is free...
Echo Nar@davidrevoy This makes me wonder how many of our current models are "designed" to intentionally misinterpret a users final request token to push for subscriptions.
@lethedata Interesting, indeed: for companies with AI/LLMs services based on prompt count or length of characters until a subscription, it sounds reasonable to think they can get extra motivation to make the user type more and turn around and around until they get a answer they consider valid or satisfactory.