Apple Stops ChatGPT Update For BlueMail On App Store. Know Everything

Apple is not allowing an update to an email app BlueMail on the App Store that uses a customised version of ChatGPT's language model, over concerns of generation of inappropriate content for kids.

AI can be misused, as previously mentioned by Mira Murati, the CTO at OpenAI.

Apple is not allowing an update to an email app BlueMail on the App Store that uses a customised version of ChatGPT's language model, over concerns about generation of inappropriate content for children, the media has reported. BlueMail app's founder Ben Volach took to Twitter to express displeasure and mentioned he disagrees with the tech giant's decision.

"Volach says Apple is un­fairly tar­get­ing Blue­Mail. The app has con­tent fil­ter­ing, and plac­ing a higher age re­stric­tion on the app could limit dis­tri­b­u­tion to po­ten­tial new users," Volach tweeted and quoted a Wall Street Journal report alongside that had his statements.

While there is no denying that AI can be misused, as previously mentioned by Mira Murati, the Chief Technology Officer (CTO) at OpenAI, the company behind AI chatbot ChatGPT, this incident is a question mark on whether language-generating AI tools, such as OpenAI's ChatGPT, are ready for widespread use.

''[AI] can be misused, or it can be used by bad actors. So, then there are questions about how you govern the use of this technology globally. How do you govern the use of AI in a way that's aligned with human values?'' Murati was recently quoted as saying in Time magazine.

"Apple has blocked the BlueMail update and continues to treat BlueMail unfairly and to discriminate against us," Volach was quoted as saying by news agency Reuters.

He also claimed that other GPT-powered apps do not seem to be restricted by Apple.

"We want fair­ness. If we're re­quired to be 17-plus, then oth­ers should also have to," Volach added in a tweet.

Earlier in December, AI research laboratory OpenAI announced ChatGPT, which is a dialogue-based AI chat interface for GPT-3 family of large language models.

The model was trained using Reinforcement Learning from Human Feedback (RLHF), using the same methods as InstructGPT, but with slight differences in the data collection setup, according to the blog post. However, the company said that the AI chat interface comes with many limitations and they plan to make regular model updates to improve in such areas.