Listen to the audio version of the article
CUPERTINO (CALIFORNIA) – It’s the day after, the magical one, for Apple. Because approximately 24 hours after the launch of Apple Intelligence, the Cupertino company’s stock soars to its highest ever, recording a tonic +5% at the opening, which brings the company well above 3 trillion in capitalisation, competing for the Wall Street leadership at Microsoft.
But how does Apple Intelligence work? And why, at the moment, does it seem like a technology capable of making Apple catch up on all the delays accumulated in recent months? Let’s go in order.
In fact, in Cupertino they have launched five new LLMs (large language models), which will interact with Siri textually and vocally, and are dedicated to translations, images, coding and more. However, they are not considered rivals to ChatGPT. Tim Cook’s company will begin implementing Apple Intelligence later this year. And the features will be available for free on iPhone 15 Pro and later versions and on all Macs and iPads in the M series. Cupertino’s AI, we imagine, will arrive in Italy no earlier than 2025 (the autumn launch is expected only in American English ), moves along two parallel lines: an AI on device, therefore extremely profiled on the user and his needs; and a more generalist AI, for which Apple relies on ChatGPT (but not only). In the first case, the possibility of being able to train an artificial intelligence with very personal data (such as health, habits, movements) but also absolutely private data seems to be the most compelling challenge. The one that the market is rewarding, and which when fully usable, will transform the smartphone into a new, intelligent and more useful device.
Not just ChatGPT
Speaking of this, in a follow-up interview with Craig Federighi, Apple’s VP and head of software, it emerged that the choice of ChatGPT is not binding. Federighi explained that the support AI will work more or less like a plugin, which users will be able to choose in the future: «Certainly – he said – people will have their preferences for certain models: maybe someone will look for an excellent model for writing creative, someone else for programming. And so we want to allow users to ultimately bring a model of their choice. This is why we imagine creating integrations with different models such as Google Gemini in the future. This is our direction.”
A moment from yesterday’s post-keynote meeting, with Craig Federighi, Justine Ezarik and John Giannandrea
Federighi also explained the decision to initially focus on ChatGPT: «it was dictated by the fact that Apple wanted to start with the best», he said.
