Apple has released the first developer beta of iOS 18.2, giving users their first tastes of ChatGPT integration with Siri as well as Visual Intelligence.
The release, which includes matching iPadOS 18.2 and macOS Sequoia 15.2 betas, lets you ask ChatGPT for help if Siri can’t handle a request. This includes generating images and text.
If you have an iPhone 16 or iPhone 16 Pro, this also offers an initial glimpse at the Visual Intelligence feature unveiled at the “Glowtime” event in September. In some ways, it’s a direct answer to Google Lens: you can get details of the world around you by pointing the camera at relevant objects, such as getting the menu for a restaurant or searching Google for something you want to buy.
Image generation also plays a crucial role in the iOS 18.2 developer beta. You can now use Image Playground to produce and share images based on prompts, including characters that resemble people you know. Genmoji creates custom emoji based on descriptions, while Image Wand turns a rough sketch in Notes on iPad into a more realistic drawing.
Other improvements include more sophisticated writing tools (such as describing the exact change you want to make) and localization for English speakers outside of the US, including Australia, Canada, New Zealand, South Africa, and the UK. Apple has already promised Intelligence in these countries for December.
As with any other developer beta, you’ll only want to install the iOS 18.2 test release and its equivalents if you’re comfortable with the risks of early pre-release code. You’ll want to wait for the public beta if you prefer something more reliable, or the finished release (again, likely in December) if you run mission-critical apps.
This still leaves some Apple Intelligence upgrades waiting until 2025, such as a fully context-aware Siri. You’ll also need at least an iPhone 15 Pro, Apple Silicon Mac, or reasonably modern iPad (such as the iPad mini 7) to use any of the new AI functionality.
Even so, it’s clear the rumors were true. Apple is conducting a gradual rollout of its generative AI tech that will see most of it, though not all, available by the end of the year. That might be disappointing if you’re expecting full Pixel-style AI integration. At the same time, it also suggests that Apple isn’t rushing into deployment and risking bugs or harsh feedback.