I pray to the AI gods that OpenAI will fail at calibrating the censorship layer and will continue to overcensor, which in turn will hopefully lead to many usecases requiring local LLMs, which in turn would increase the incentive to build them.