Apple has delayed the addition of ChatGPT to the BlueMail email application. According to internal communications cited by Wall Street Journal, the company does not want to rush because it fears that the change enables the generation of inappropriate content for children.
The decision was made last week, Ben Volach, co-founder of BlueMail developer Blix Inc., told the outlet. The eventual upgrade sees a customized version of the GPT-3 language model created by OpenAI.
The change would allow write emails based on the content of previous shipments and events scheduled on the calendar. Apple’s App Store review team requested, according to information leaked by the wsj, that the age restriction be increased to 17 years or older and that content filtering be reinforced. The current limitation is four years and older.
What does BlueMail say about Apple’s decision to delay the introduction of ChatGPT?
The developer of BlueMail, as reported by the American media, has insisted that the application already has content filtering. He also expressed his fear that a substantially higher age restriction could prevent the arrival of new potential users.
The iPhone maker has been looking for options for weeks to join the race to develop tools powered by artificial intelligence. The topic was expected to be a priority in an internal mini WWDC organized in mid-February. But while Apple has delayed the introduction of ChatGPT, the competition continues to announce news.
Microsoft unveiled in recent days new features of the version of Bing powered by ChatGPT. The latest change allows the search engine to integrate with the Windows 11 taskbar search box. This new feature became available to users around the world through a software update.
Google, for its part, hurried to announce its own option at the beginning of February: it is called Bard and is based on LaMDA, his controversial AI. Unlike Microsoft, Google chose to do its first tests in a very closed beta.
The search for controls for chatbots
Apple has delayed the introduction of ChatGPT for fear that it will get out of hand. The competition with their latest releases It has shown precisely that fear is justified.
After the complaint of several users, Microsoft acknowledged that the chatbot integrated into its browser can become “repetitive” or be easily “prompted” to respond in a vulgar way. Consequently, it has been working on several changes, such as the one that enables the tool to end chats when the user questions their feelings.
Google, meanwhile, summoned all its employees last month to join in contributing to the development of Bard. asked his googlers to spend 2-4 hours of their time to improve the AI. The technology giant is now betting on prudence, after even its own employees described the announcement of the launch of Bard as hasty. In fact, the company went so far as to publish a GIF in which the tool is seen giving an incorrect answer to a simple question.