OpenAI needs people to stay stuck inside the app because they are gearing up for ads. Dark patterns are escalating everywhere. Duolingo is nowadays also becoming like a blackmailer: "No one ignores me (the silly owl) this long and doesn't regret it later" or a "broken app icon" that can only be fixed by doing a specific task.
It's back to the original GPT-5 state that everyone hated. Everyone's not just unsubscribing to ChatGPT for political reasons; I believe the tone also puts everyone off.
I've noticed this too. "...two small tweaks" is just a variation on the "one weird trick" you see in chumboxes[0]. I guess this sort of enshittification was inevitable.
This is the real cost of an ad driven model.
I’m sure a key KPI is engagement.
Maybe they are prepare for ad?
Wrong post, wrong product. This thread is about GPT fear-mongering, not Claude Code workflows.