Apple’s top software executives decided early last year that Siri, the company’s virtual assistant, needed a brain transplant.
The decision came after administrators Craig Federighi and John Giannandrea tested OpenAI’s new ChatGPT chatbot. The product’s use of generative artificial intelligence, which can write poetry, generate computer code and answer complex questions, has made Siri look dated, said two people familiar with the company’s work who were not authorized to speak publicly.
Introduced as the original virtual assistant on every iPhone in 2011, Siri has been limited to individual requests for years and has never been able to follow a conversation. He often misunderstood the questions. ChatGPT, on the other hand, knew that if someone asked about the weather in San Francisco and then asked, “How about New York?” that user wanted another prediction.
The realization that the new technology was leapfrogging Siri set in motion the tech giant’s most significant reorganization in more than a decade. Determined to catch up in the tech industry’s AI race, Apple has made generative AI a tentpole project — a special, internal label the company uses to organize its employees around once-a-decade initiatives.
At its annual developer conference on June 10, Apple is expected to show off its AI work, which will introduce an improved Siri that is more conversational and versatile. Siri’s core technology will include a new generative artificial intelligence system that will allow her to chat instead of answering questions one-by-one.
The Siri update is at the forefront of Apple’s broader efforts to incorporate generative artificial intelligence into its business. The company is also increasing the memory in this year’s iPhones to support the new Siri capabilities. And it has discussed licensing complementary AI models that power chatbots from several companies, including Google, Cohere and OpenAI.
An Apple spokesman declined to comment.
Apple executives worry that the new artificial intelligence technology threatens the company’s dominance of the global smartphone market because it has the potential to become the main operating system that replaces the iPhone’s iOS software, two people not familiar with Apple management’s thinking said. permission to speak publicly. This new technology could also spawn an ecosystem of artificial intelligence apps known as agents that can order Ubers or schedule appointments, undermining Apple’s App Store, which generates about $24 billion in annual sales.
Apple also fears that if it fails to develop its own AI system, the iPhone could become a “dumb brick” compared to other technologies. While it’s unclear how many people regularly use Siri, the iPhone currently takes in 85 percent of global smartphone revenue, bringing in more than $200 billion in sales.
That sense of urgency helped fuel Apple’s decision to scrap its other big bet — a $10 billion project to develop a self-driving car — and reassign hundreds of engineers to work on artificial intelligence.
Apple has also explored building servers powered by iPhone and Mac processors, two of the people said. Doing so could help Apple save money and create consistency between the tools used for processes in the cloud and on its devices.
Instead of directly competing with ChatGPT by releasing a chatbot that does things like write poetry, three people familiar with its work say Apple is focusing on making Siri better at managing tasks it already does, including setting timers, creating calendar appointments and adding apps. directed to do. Add items to your grocery list. It can also summarize text messages.
Apple plans to make the improved Siri more private than rival AI services because it will process queries on iPhones rather than remotely in data centers. The strategy will also save money. OpenAI spends about 12 cents for about 1000 words generated by ChatGPT due to cloud computing costs.
(The New York Times sued OpenAI and its partner Microsoft in December for copyright infringement of news content about AI systems.)
But Apple faces risks by relying on a smaller AI system embedded in iPhones rather than a larger system housed in a data center. Research shows that small AI systems are more likely to make errors known as hallucinations than large systems.
“Having a conversational interface that understands language and context has always been the vision for Siri, but it’s a tough challenge,” said Siri co-founder Tom Gruber, who worked at Apple until 2018. “Now that technology has changed, it should be possible to do better than that. As long as responding to something is not a one-time effort, then they should be able to avoid the problem.”
Apple has several advantages in the AI race, including more than two billion devices in use worldwide to deploy its AI products. It also has a leading semiconductor team that develops sophisticated chips capable of performing artificial intelligence tasks such as facial recognition.
But over the past decade, Apple has struggled to develop a comprehensive AI strategy, and hasn’t made much progress since the introduction of Siri. The assistant’s struggles dampened the appeal of the company’s HomePod smart speaker, as it failed to consistently perform simple tasks like performing a song request.
John Burkey, who worked on Siri for two years before founding Brighten.ai, a generative artificial intelligence platform, said the Siri team didn’t get the attention and resources that other groups within Apple were given. The company’s divisions, such as software and hardware, operate independently of each other and share limited information. But AI needs to go through products to succeed.
“It’s not in Apple’s DNA,” Mr. Burkey said. “It’s a blind spot.”
Apple has also struggled to recruit and retain top AI researchers. Over the years, it has acquired AI companies led by leaders in the field, but they all left after a few years.
Their reasons for leaving are varied, but one factor is Apple’s privacy. The company publishes fewer AI-related papers than Google, Meta, and Microsoft, and it doesn’t attend conferences like its rivals.
“Research scientists are saying, ‘What are my other options?’ Can I go back to academia? Can I go to a research institute, work somewhere, a little more open?’” said Ruslan Salahutdinov, a leading artificial intelligence researcher who left Apple in 2020 to return to Carnegie Mellon University.
In recent months, Apple has increased the number of AI documents it publishes. But prominent AI researchers have questioned the value of the filings, which are more about creating a meaningful business impression than providing examples of what Apple might bring to market.
Tsu-Jui Fu, an Apple intern and PhD student in artificial intelligence at the University of California, Santa Barbara, wrote one of them. Apple’s latest AI documents. Last summer, he developed a system for editing photos with written commands rather than Photoshop tools. He said that Apple supported the project by providing him with the necessary GPUs to train the system, but that he had nothing to do with the AI team working on Apple products.
Although he says he’s interviewed for full-time jobs at Adobe and Nvidia, he plans to return to Apple after graduation because he thinks he can make a bigger difference there.
“Apple’s AI product and research is nascent, but most companies are very mature,” Mr. Fu said in an interview with The Times. “Instead of being part of a team doing something at Apple, I can have more room to lead a project.”