Apple Intelligence now allows users to ask the system to rewrite a sentence, summarize messages or emails, generate pictures, and even emoji. Siri has also been improved with AI, so it now understands the context of what’s on the screen and allows users to control more aspects of their device.
Apple uses both on-device and cloud computing for its Intelligence features.The company has also collaborated to integrate ChatGPT with Siri. Although Apple has devoted a lot of effort into developing its own language models, Federighi recognizes that there are other strong LLMs available.
“These very large frontier models have interesting capabilities that some users appreciate, and we saw that integration into our experiences could make [those capabilities] much more accessible than [they are] today,” the manager stated. According to Federighi, GPT-4o is one of the best LLMs available today, which is why Apple wants to integrate it into Siri.
To make things clearer for consumers, Siri will ask for permission to send queries to ChatGPT whenever Apple Intelligence is unable to answer a question. Federighi also stated that Apple intends to introduce support for more third-party language models in the future, allowing customers to select what works best for them.
China is one of Apple’s most important and highly regulated markets. These laws make it far more difficult for non-Chinese companies to sell AI models in China, but what does this mean for Apple? Without going into too much detail, Federighi said that the business is working to “find a way” to bring Apple Intelligence to China.
“We don’t have timing to announce right now, but it’s certainly something we want to do.” Apple Intelligence will first be available in US English, although it is unknown whether customers in other regions will be able to use it.
A portion of the requests provided to Apple Intelligence are processed by Apple’s AI servers known as Private Cloud Compute, or PCC. The business has devised a method of storing data requests and cryptographically destroying them once they have been completed, so Apple sees nothing.
Although Federighi thinks that future chips will be capable of running larger language models, he emphasizes the importance of having an online model for giving up-to-date information.
“I couldn’t rule it out,” he says, “but even in that world, I think you would expect your device to reach out to knowledge stores that are outside the device, in order to service your request.”
So even in the future, I believe there will be a place for contacting foreign services.”
Read More: Click Here