When Apple Intelligence was revealed at the Worldwide Developers Conference (WWDC) earlier in June, Apple made it clear that not every feature of its artificial intelligence (AI) would be ready when iOS 18 and the company’s other new operating systems launch in the fall. But now it looks like we might have to wait until 2025 before every aspect of Apple Intelligence sees the light of day.
According to a new Power On newsletter from Bloomberg journalist Mark Gurman, Apple Intelligence will not be available for developer testing until “later this summer,” despite Apple having already launched developer betas of its upcoming operating systems, including iOS 18 and macOS Sequoia.
As well as that, when these systems do launch, Apple Intelligence will only be a preview of what is to come, with language support limited to American English and not all Apple devices being compatible at first.
Gurman claims that many features won’t launch until 2025. That includes Siri acquiring contextual knowledge of what you’re doing at any given moment and using that to help you find things and perform tasks. What’s more, Siri won’t gain more precise control over apps and your device – such as editing a photo and then sending it as an email attachment – this year either, Gurman claims.
Some features will arrive in 2024, the report says, but not in time for the launch of the new operating systems. New features in the Mail app, AI-based code writing in Xcode, and potentially even Apple’s headline partnership with ChatGPT could all miss that deadline.
Better safe than sorry
While this might sound like disappointing news, there are actually several good reasons why Apple is taking this approach. For one thing, Gurman believes that the company is seeking to avoid the mistakes made by its rivals in the AI arena. ChatGPT developer OpenAI has been plagued by the tendency of its chatbot to churn out false and misleading information, while Microsoft has had to delay its AI-powered Recall feature after a sizeable backlash over its privacy-violating abilities.
A slower rollout gives Apple the opportunity to train its AI on other languages, which is a very demanding undertaking. It also allows more time for Apple to build out the cloud infrastructure that will power things like Private Cloud Compute, while the company will be able to manage its workers better and assign them to new tasks as previous ones get completed, rather than facing “chokepoints with staffing,” Gurman says.
Apple has traditionally preferred to take its time and get new products right, rather than rush something unfinished out of the door that could negatively impact its customers and its reputation. Given the dangers of getting AI wrong, this slowly-slowly approach may very well turn out to be the right one.
