Apple would be two years behind on artificial intelligence

Last June, Apple unveiled Apple Intelligence functions during WWDC 2024, revealing the firm's plans for artificial intelligence. We were then able to see numerous additions and new functions linked to AI, but all this also set the firm's strategy in stone: its goal, unlike Microsoft, Google or Meta, is not to offer its own generative artificial intelligence, but indeed to use that of others.

“2 years behind in generative artificial intelligence compared to industry leaders”

Apple Intelligence mainly relies on models that can run on the device, which means that it is necessary to have an A17 or M1 chip (or later) with 8 GB of RAM to benefit from it. The fact that Apple Intelligence operates directly locally thus limits the volume of information, which explains why Apple has at the same time announced a partnership with OpenAI to integrate ChatGPT into the entire system, for users who wish it. This ChatGPT-4o integration hides a bigger problem: According to Gurman, Apple's in-house artificial intelligence is years behind.

The journalist therefore suggests that ChatGPT-4o would offer answers 25% more precise than Siri, and can also answer 30% more questions compared to Apple's assistant. Then he adds: “some Apple employees consider Apple's generative artificial intelligence technology to be at least two years behind industry leaders“. So what is Apple's strategy to fill this gap? According to Gurman, by 20206, Apple Intelligence should run on all Apple devices with a screen, including iPad, iPhone SE and less expensive models. on the other hand, the functions of the service should then be much more advanced. It remains to be determined how these features will be integrated, and whether they will be well accepted by users. As a reminder, with us, Apple Intelligence won't be available until 2025.