Apple’s new AI features, introduced with iOS 18.2, enhance capabilities like Siri and Visual Intelligence, but face limitations in France until April 2025. Users can experience advanced functionalities, including integration with ChatGPT, but must navigate regional restrictions and potential performance issues. Siri’s improvements allow for better comprehension of complex inquiries, yet it often fails to utilize ChatGPT effectively. Visual Intelligence offers innovative features, though it suffers from touch sensitivity. Consolidating these tools could enhance user experience.
Imagine having an AI assistant like Siri that can tackle complex inquiries, a camera equipped with recognition capabilities, or a tool to enhance your spelling. On the surface, Apple Intelligence appears to offer a variety of incredibly useful features. With the launch of iOS 18.2, which incorporates ChatGPT into the iPhone, Apple’s AI functionalities have become even more captivating.
In France, however, Apple Intelligence won’t be available until April 2025, and it may come with certain limitations (Apple has been vague about this, but we can speculate that Siri’s performance in French might be subpar). While it’s possible to access it from the EU ahead of time, doing so would mean losing access to your French App Store, Apple Music, Apple TV+, and Apple Fitness+ accounts. This frustrating trade-off prevents many Europeans from experiencing this exciting new feature from Apple.
Apple Intelligence: Automatic Activation Upon Arrival
One notable aspect is the sophisticated locking mechanism that Apple has in place for its services within the European Union.
While in France and onboard a flight, the iPhone indicated that the service was “not available in our region.” Yet, just moments after landing in the United States, the service became accessible. Aside from GPS, which can be manipulated with the right tools, it seems Apple utilizes cell towers to restrict its services to the U.S. This level of restriction is rare; typically, a VPN suffices to bypass such geo-restrictions.
Another intriguing finding is that Apple Intelligence activates automatically, significantly altering various features of iOS. To our knowledge, this marks the first instance where Apple provides two distinct user experiences within the same operating system. In the long run, it’s hard to envision Apple Intelligence remaining a feature that users must manually toggle. Additionally, activating the service requires a download of 3.18 GB with iOS 18.2, which we suspect will eventually be integrated into the operating system itself.
Siri, ChatGPT, and Visual Intelligence: A Lack of Clarity
Challenges with Siri
Upon activation of Apple Intelligence, users are greeted with a brief tutorial. This introduction showcases the revamped Siri (featuring a lively animation and a text-based question system) alongside a summary of notifications. Despite this, Apple Intelligence may not activate successfully on the first attempt, though a simple restart can resolve the issue. Users can also manually enable ChatGPT within the settings, allowing Siri to request assistance from OpenAI’s servers when it lacks an answer. By default, Siri relies solely on its own capabilities.
What’s new with Siri, you might ask? Its primary enhancement is the ability to process lengthy questions, where the prior version would have simply replied, “I didn’t understand.” The updated Siri can grasp the intent behind poorly phrased requests, opting to respond from its database or consult ChatGPT when necessary.
However, Apple seems to be missing the mark in this area. Frequently, for reasons unknown, Siri opts to inform users that it cannot answer a question instead of quickly checking with ChatGPT. Users must then manually rephrase their inquiry to start with “Ask ChatGPT” to obtain a comprehensive response.
Moreover, there are instances where Siri unnecessarily directs simple queries to ChatGPT, such as asking for travel times to locations. The downside? ChatGPT lacks access to the user’s location, rendering it unable to provide accurate answers. For example, when asking about the duration to reach the Nintendo World store, Siri forwarded the question to ChatGPT, which returned a generic response.
While on the go, we often attempted to engage Siri using our AirPods to ask about nearby monuments. In almost every instance, Siri failed to comprehend the question and neglected to check with ChatGPT. Users must explicitly state “Ask ChatGPT” to receive an answer, diminishing the service’s appeal. What’s the point of using Siri when the ChatGPT app offers a superior voice interaction experience?
For instance, when asked, “What is the name of the place with the carousel and the beautiful view in New York?” ChatGPT Voice responds with “Jane Carousel” and suggests picturesque photo spots, while Siri simply recites the Wikipedia entry for Central Park.
Visual Intelligence: A User-Friendly Interface
Another feature partially built upon ChatGPT is Visual Intelligence. This application is exclusive to the iPhone 16 and can only be accessed through the Camera Control button. Inside, users can find an AI-enhanced camera that allows for text translation, summarization, Google Lens comparisons, and even questions directed at ChatGPT.
However, Visual Intelligence has a significant drawback: it is overly sensitive to touch. An accidental tap on the screen can reset the conversation, forcing users to start anew. Aside from this, it stands as one of the most impressive features of Apple Intelligence. The ability to photograph a statue and inquire about its significance is remarkable, as is the convenience of scanning a foreign menu to identify vegetarian options.
To enhance Visual Intelligence, Apple could consider three key improvements:
Despite these minor issues, the service largely fulfills its promises. It’s unfortunate that Apple hasn’t utilized its own models, but ChatGPT performs adequately. An added benefit is that ChatGPT can communicate in French, enabling users to pose questions in Molière’s language.
Nevertheless, a significant question remains: why integrate a service when existing apps like ChatGPT or Google Lens can perform similar tasks?
Apple Should Unify Apple Intelligence into a Single Interface
The fundamental issue with Apple Intelligence in its current version is clear: Apple has yet to fully embrace its transition into the realm of artificial intelligence.
It seems impractical to have multiple interfaces for querying your iPhone. Siri (voice and text), Spotlight (local search), Visual Intelligence (image recognition), Look Up (photo analysis)… All of these features should be consolidated, potentially under the Siri brand. We envision the ability to send a screenshot to Siri for answers or verbally instruct Siri to “take a photo and tell me what you see.” The coexistence of various functionalities complicates the user experience, especially when an application like ChatGPT can seamlessly handle everything. The silver lining is that Siri is expected to improve with iOS 18.4 in April, paving the way for enhanced multimodal capabilities.
Genmoji and Notifications: Highlights Amid Room for Improvement
Within its Apple Intelligence package, the tech giant from California includes a variety of services. Among these features, there’s a tool for generating video memories, a magic eraser to eliminate unwanted elements from photos, and a system for summarizing emails or web pages.