Apple (AAPL) announced its long-awaited generative AI initiative called Apple Intelligence on Monday at its WWDC conference in Cupertino, Calif. The technology, which marks Apple's first step toward generative AI, will be deeply integrated across the company's hardware and software products, from iPhones and Macs to Mail, Messages and Photos.
Rather than a broad AI system like ChatGPT or Google’s AI Overview, Apple is positioning Apple Intelligence as a unique service that can understand you and your data.
Apple Intelligence will be available on iPhone 15 Pro, iPads, and Macs with Apple's M1-series chips and later this fall.
The biggest change is coming to Apple's Siri. The original smartphone voice assistant has been in desperate need of a refresh for years, and Apple Intelligence will give it just that: The company says the assistant will have a new look, feel more natural, and be more responsive.
Like other generative AI-powered assistants, you can ask follow-up questions or interrupt your request during the request. Siri can now respond to your requests with typed text if you don't want to request it out loud. You can also ask Siri to make your request using ChatGPT instead of Apple's proprietary model.
Apple says the updated version of Siri is more aware of the context of your Apple products and can give you accurate answers when you ask it about how different features and settings work. On-screen recognition allows Siri to understand what's on your screen and take action on it. So if a friend sends you an address in a message, you can have Siri save it.
Apple has been particularly strong in strengthening Siri's ability to understand your own data. For example, you can ask a question like, “Show me pictures of Stacey in her pink coat in New York,” and the assistant will give you exactly the photo you're looking for from the Photos app. You can then tell Siri to move the photo to another app, like Mail in Mail. These requests also work across third-party apps.
If you need to pick up someone at the airport, you can ask Siri when they'll be arriving, and Siri will look up the flight information they previously shared via email and check real-time flight tracking data to determine when you should leave.
Beyond Siri, the device can now prioritize notifications to surface the most important notes and minimize the less important ones. Writing tools allow you to rewrite, write, or summarize information, which is automatically available in Notes, Mail, and a host of third-party apps. Across the app, you can create AI-generated images of people, places, and animals in three different styles.
Apple says that while many of its generative AI models run on-device, some require access to the cloud. But Apple has traditionally avoided forcing users to use cloud-based services when it comes to personal data, so the company developed a new cloud service called “Private Cloud Compute.”
The private cloud computing servers are custom built using Apple Silicon with privacy features built in. When you make a request, Apple Intelligence determines whether it can complete the request using on-device processing or if it needs to connect to the cloud. However, Apple says it doesn't store any of your information in the cloud that's used to complete Apple Intelligence requests.
Wall Street has been eagerly awaiting Apple to debut its generation capabilities ever since Microsoft (MSFT) unveiled its OpenAI-powered Bing chatbot (now named Copilot) in 2023. Google (GOOG, GOOGL) soon followed Microsoft's lead with the launch of its Bard chatbot (later renamed Gemini). Social media giant Meta (META) is also incorporating AI into its platform offerings through its behind-the-scenes recommendation software and Meta AI chatbot.
But these companies have also struggled with their AI deployments. Google's Bird famously debuted with an ad that gave wrong answers to prompts, and the company's Gemini image-generating app showed historically inaccurate images of people from different time periods. More recently, the rollout of a massive AI overview, a major effort aimed at transforming the company's vital search platform, drew derision when it produced results telling people it's safe to eat stones or put glue on pizza.
Meanwhile, ahead of its Build conference, Microsoft announced its Copilot+PC standard for laptops and desktops, as well as a new Recall app for Windows 11. The software takes screenshots of just about anything you use on your computer, helping you recall apps and information you've previously used. But the app was slammed as a potential security nightmare, and Microsoft had to update the software before its general release to make it opt-in instead of on by default, tightening up security.
Meanwhile, Meta faced heavy criticism after it prevented consumers from turning off the Meta AI chatbot within its apps.
While these issues won't slow the companies' performance, Apple is an entirely different breed: The iPhone maker has built its reputation on security and software that works, for the most part, out of the box. If its AI-generating efforts suffer from the same flaws as Google and Microsoft, it could hurt the brand's reputation.
Moreover, Apple's misstep could tarnish generative AI's image more broadly: after all, if three of the world's largest tech companies can't roll out a new kind of software without major issues, can users really trust generative AI?
Apple still has time to work on Apple Intelligence before it hits users' devices in September of this year.
Contact Daniel Howley at dhowley@yahoofinance.com. Follow him on Twitter. Daniel Howley.
Click here for the latest technology news impacting the stock market.
Read the latest financial and business news from Yahoo Finance
