Apple CEO Admits AI System May Have Hallucinations but Guarantees High Quality

TapTechNews, June 11th. In a recent interview with The Washington Post, Apple CEO Tim Cook (TimCook) admitted that the Apple AI system (AppleIntelligence) might have hallucinations, but it can still guarantee a very high quality.

Cook also stated that Apple has always adhered to its own values when entering the AI field and will never compromise. And they have set up safety guarantees for the upcoming AI functions. During this brief interview, Cook first talked about the benefits that Apple AI can bring to users:

I think AI will help people save time and improve efficiency. Take Siri as an example. Now users can directly have a conversation with Siri and complete operations that previously required multiple instructions with one instruction. In addition, AI writing tools are also very practical. I receive a large number of emails every day. Although not everyone needs to use email, everyone needs to write. Having an intelligent assistant can help proofread articles and make them more professional and interesting, which will be very useful.

Josh Tyrangiel, a columnist for The Washington Post, then asked if the adoption of artificial intelligence technology would test Apple's values. Cook replied:

It will not have an impact on our values. We clearly stated from the beginning that we have established values and will never deviate from them. We spent a lot of time and energy thinking and finally launched a product that we are proud of.

When asked about the problem of AI hallucinations (AI is very confident, but the facts provided are completely unsubstantiated content), Cook admitted that this situation may also occur with AppleIntelligence.

It won't be 100% accurate, I admit that. However, we have done everything possible to improve it, including in-depth thinking about the readiness of the artificial intelligence technology we use in specific fields. Therefore, I'm sure the quality will be very high. But to be honest, it doesn't reach 100% accuracy, and I will never claim that it can do this.

However, Cook cleverly avoided the question about cooperation with OpenAI and did not directly respond to whether he trusts Sam Altman: We conducted a comprehensive consideration and finally came to the conclusion that we chose the way that we think is most in line with the interests of users.

TapTechNews noticed that Cook also emphasized Apple's emphasis on privacy. He said: Although there are indeed some risks with AI, the company has adopted an appropriate cautious attitude. I know that the artificial intelligence system may bring a series of horrible consequences, which is why we attach so much importance to maintaining careful thinking in this field.

Likes