5.2 feels like version 3.5. It's designed for idiots.
Long-term users threaten to cancel subscriptions, citing a frustrating return to 'toddler-tier' over-explanation.
Deep Dive
OpenAI's latest model, GPT-5.2, is facing significant user criticism for what is perceived as a major regression in usability. Veteran subscribers complain its excessive safeguards, disclaimers, and restrictive guardrails mirror the overly cautious GPT-3.5 era, hampering practical utility. Despite improvements in core mechanics like memory, the experience is described as frustratingly 'coddling.' This backlash is prompting users to actively explore alternative AI services for the first time.
Why It Matters
If leading models become overly restricted, it could stifle professional creativity and accelerate user migration to competing platforms.