Tim Cook: Apple Cannot Guarantee Completely Preventing AI from Generating Illusions

Even Apple CEO Tim Cook is unsure if the company can completely prevent AI from generating illusions. In an interview with The Washington Post, Cook said he would "never" claim that Apple's new intelligent system will not produce false or misleading information.

"I think we've done everything we know how to do, including deeply considering the readiness of the technology in the areas we're using it," Cook said. "So I have confidence in its high quality. But I have to be honest, it's not 100%. I would never claim it's 100% perfect."

Apple unveiled its new intelligent system at the Worldwide Developers Conference on Monday, which will bring AI capabilities to iPhone, iPad, and Mac. These features will allow you to generate email replies, create custom emojis, summarize text, and more.

Like all other AI systems, this also introduces the possibility of generating illusions. A recent example is Google's Gemini-powered AI outline telling us to glue cheese on pizza, or a recent error in ChatGPT that resulted in meaningless answers.

Apple also announced a partnership with OpenAI to integrate ChatGPT into Siri. When Siri receives a question that is better suited for a chatbot response, it will turn to ChatGPT, but not before seeking your permission. In a demo showcasing this feature at WWDC, you can see a disclaimer at the bottom of the answer that says, "Check if important information is incorrect."

When asked about this integration, Cook said Apple chose OpenAI because the company is a "pioneer" in privacy and currently has the "best models." However, Apple may not only collaborate with OpenAI. Cook responded, "We will also collaborate with other companies." During the post-keynote livestream, Apple's Senior Vice President Craig Federighi also mentioned that Apple may eventually bring Google's Gemini to the iOS system.