Apple enters the race for generative AI


Following this week’s announcements of new generative artificial intelligence projects at OpenAI and Google, all eyes in the industry have been on Apple. Tim Cook, the CEO of the American company, has just announced that the company will soon unveil its next steps in the new field of open AI, which has recently exploded with the introduction of ChatGPT. This will take place on June 10, at the opening of the annual Worldwide Developers Conference, when the new features of iOS 18, the latest operating system for the iPhone, will be announced. In a general lack of details, multiple rumors have circulated in recent weeks regarding potential directions in which the tech giant could move.

Analysts appear to agree that Apple’s entry into the generative AI race will not be aimed at directly competing with OpenAI and Google, but rather that the iPhone maker could break the current link between the two choosing one of them to work with its chatbots and other AI services. According to the latest leaks, Apple could be close to signing a deal with ChatGPT to integrate it into the iPhone: although the same sources note that discussions are still underway with Google regarding the use of its Gemini technology. For both competitors, having such privileged access to more than a billion iPhone users would be a key objective of their fierce rivalry.

Since the launch of the first iPhone in 2007, Google has paid Apple a hefty sum – $20 billion in 2022 alone – to make its search engine the default on phones. While it’s not yet clear whether Apple will go with OpenAI or Google, and to what extent it will incorporate AI capabilities, the company’s top policy analysts believe it will move toward some type of contract different from that of the search engine.

Namely, instead of setting ChatGPT or Gemini chatbots as default on the iPhone, the deal could involve using their technology. According to BloombergMark Gurman of: “The two companies are in active negotiations to allow Apple to license Gemini, Google’s set of generative AI models, to power some new features coming to iPhone software this year. ” In its daily newsletter Strategy, Ben Thompson writes that he thinks Google has a better infrastructure than OpenAI when it comes to being able to handle the avalanche of traffic and data processing that iOS integration would require. Meanwhile, John Gruber, in his technology-focused blog Bold Fireballbelieves that: “I could also see Apple negotiating deals to use multiple AI vendors behind the scenes, treating them as white-label providers, while presenting the features to users under the Siri brand. »

Many expectations for Apple’s next moves focus on what could be a radical overhaul of its Siri voice assistant to make it truly intelligent. Since Siri’s debut in 2011, a day before Steve Jobs’ death, it has seen few significant updates and while it can solve simple tasks like calling a contact from its phone book, its voice recognition system still resists other requests, like playing a certain song or album. Google’s Assistant and Amazon’s Alexa overtook Siri years ago and are considered more efficient and smarter.

The fall of Siri

The arrival in 2022 of ChatGPT, with its vastly superior ability to listen, give answers and even chat with users, has made it all the more obvious how far Apple lags behind the competition when it comes to command-based assistants. voice. And this week, the OpenAI demonstration of its new voice conversation mode – reminiscent of the film Her in that it’s able to tell a story on any topic you ask, give it the user’s favorite dramatic twist, and interpret it in multiple languages ​​- sent Siri straight to the box forgotten toys.

If Apple chooses to use ChatGPT or Gemini to speak directly to iPhone users and respond to their requests, it would be the end of Siri. And Apple would lose considerable control over privacy, which has become one of its biggest talking points in recent years. Amid widespread sentiment that Apple is lagging behind in artificial intelligence, Cook said in a recent quarterly presentation: “We believe we have advantages that will differentiate us in this new era, including Apple’s unique combination of hardware, software and service integration, revolutionary Apple silicon with our cutting-edge neural engines and our unwavering focus on privacy that underpins everything we create.

According to John Gruber, who has long accurately predicted new Apple announcements, “Apple’s own efforts in language models (which are driven by generative AI) appear to be geared toward processing in its own device ”, leveraging the capabilities of its own processors and retaining full control over privacy, compared to services like ChatGPT, which use the enormous processing capabilities of cloud server farms. Apple may choose to create relatively simple AI mechanisms for iPhones with iOS 18 software, which will debut in September. They would handle sensitive information, such as summarizing an audio message or writing a response to an email, in line with what Samsung has already integrated into its latest cell phones.

Apple’s privacy and its own chips

Compared to its competitors, Apple has a distinct advantage in that it makes its own processors for its phones, tablets and computers. Last year, the technology industry was surprised by how quickly it evolved: in just six months, the company went from launching the M3 processor for its MacBook Pro to launching the next generation M4 for the new iPad Pro, which just went on sale. Since 2017, Apple has included in its processors a neural core exclusively dedicated to performing artificial intelligence tasks on its devices; the most recent leaks indicate that the tech giant is preparing these chips for use by its own cloud servers.

“We remain very optimistic about our opportunity in the generative AI space. We are making significant investments and we look forward to sharing some very exciting things with our customers soon,” Cook added in his recent speech to analysts and investors. Until the presentation on June 10, with which Apple will launch its annual meeting with developers of applications for its devices, the company will not make its plans public. And, besides the leaks, it will not be possible to deduce many concrete details from the acquisitions of AI activities that the company has made over the past year, nor from the scientific articles on the language models that it published. Neither were considered very relevant by experts.

The only thing Apple let slip, although it’s far from a big revelation, is that it wants to make a big entry into this hot new area. “We are spending a tremendous amount of time and effort on artificial intelligence,” Cook told his investors in February. Since then, its marketing team has worked to highlight its devices’ capabilities in this area, making hard-to-substantiate claims such as “customers love the incredible AI performance of the latest MacBook Air and MacBook Pro models.” .

While we await the real details on how Apple will make its promised entry into the generative AI race, the company appears to have the resources necessary for a significant launch in the field, following the cancellation of its planned several billion dollars. to create an autonomous car, and the recent completion of its first mixed reality glasses, the Vision Pro. AI is a marathon, not a sprint, and we’re still in its early days with the new ChatGPT voice assistant and the recently introduced Google AI search engine update – but these projects indicate exactly what’s at stake when It’s about creating a new way of interacting with our computers, mobile phones and tablets.

Register for our weekly newsletter to get more English media coverage from EL PAÍS USA Edition



Source link

Leave a Comment

Your email address will not be published. Required fields are marked *

Scroll to Top