Apple could only bring on-device AI features to iPhones with iOS 18 According to a report from Mark Gurman (via 9To5Mac), AI features on iPhones will be powered entirely by a Large Language Model (LLM) developed by Apple, and all the AI processing will happen on-device.
We can expect Apple to heavily market the privacy and speed benefits of using on-device AI processing.
These AI features will reportedly debut with iOS 18 and will be showcased during WWDC 2024.
While on-device AI has privacy and speed benefits, it isn’t as powerful as AI processing offered by dedicated servers in the cloud.
However, it is possible that Apple will only introduce AI features that work well with on-device AI processing.
This differs from Samsung’s approach of using a mix of on-device and cloud-powered AI features.
As seen on the Galaxy S24 and older phones that have received One UI 6.1 with AI features, some AI features work fast, while others are slow, depending on the workload and complexity.
A few weeks ago, it was revealed that Apple was in talks with Google about using Gemini to power some AI features in iOS 18.
With the Galaxy S24, Samsung debuted a number of AI-powered features. The Galaxy S24 was dubbed Samsung’s first artificial intelligence phone. With the rise of AI, it is the first smartphone brand to take a major risk. Apple, like its South Korean rival, will take a slightly different approach, but it will still follow Samsung in due course.
iPhones running iOS 18 and later could only get on-device AI features from Apple.
A Large Language Model (LLM) created by Apple will power all AI features on iPhones, according to a report from Mark Gurman (via 9To5Mac). All AI processing will take place on the device itself. Apple is likely to aggressively promote the speed and privacy advantages of on-device AI processing. It is reported that these AI features will make their appearance at WWDC 2024 and launch with iOS 18.
While on-device AI offers advantages in terms of speed and privacy, it lacks the processing power of cloud-based dedicated servers for AI. Apple might, however, limit the AI features it releases to those that are compatible with on-device AI processing. Better auto-replies and Siri requests, for instance, might be available.
This is different from Samsung’s strategy, which combines AI capabilities powered by both the cloud and on-device. For AI processing, Samsung combines Google’s Gemini with its own LLM. Using a straightforward toggle, users can choose to process AI data locally on the gadget. As evidenced by the Galaxy S24 and other older phones running One UI 6.1 with AI features, the workload and complexity of each AI feature determines how quickly or slowly it operates.
Apple has not disclosed any AI features that may launch with iOS 18, but they could include enhanced image editing capabilities, more dependable autocorrection when typing, and improved language translation. It was made public a few weeks ago that Apple and Google were in discussions about utilizing Gemini to power certain AI features in iOS 18. It’s unclear, though, if that agreement has been reached.
