TL;DR:
- Apple has significantly increased its investment in artificial intelligence, with a focus on conversational AI.
- A specialized “Foundational Models” team of 16 individuals is dedicated to developing language models.
- Apple’s daily expenditure on training language models reaches millions of dollars.
- The company aims to automate multi-step tasks through voice assistants like Siri, potentially introducing this feature in iOS 18.
- Apple is also exploring AI applications for generating videos, images, and multimodal AI.
- The Ajax chatbot, with 200 billion parameters, is positioned as a contender, although newer OpenAI models outpace it.
Main AI News:
In a recent report from The Information, it has come to light that Apple has significantly bolstered its financial commitment to the realm of artificial intelligence. This surge in investment underscores the tech giant’s dedication to advancing its capabilities in AI and machine learning.
Despite initial skepticism from Apple’s AI head, John Giannandrea, regarding AI chatbots, he initiated the formation of a specialized team focused on conversational AI approximately four years ago. Speculation about Apple’s venture into the domain of AI chatbots, often referred to as “Apple GPT,” had previously surfaced through insights shared by Bloomberg’s Mark Gurman. In July, Gurman hinted at Apple’s exploration of expansive language models, with certain Apple staff gaining access to an internal chatbot known as “Ajax.”
The emergence of OpenAI’s ChatGPT in 2022 created a newfound demand for chatbots within the tech industry. Major players like Microsoft and Google swiftly launched their own iterations, but as of now, there’s no indication of Apple introducing a consumer-centric product in this category anytime soon.
Apple’s specialized team, known as the “Foundational Models” team, focusing on conversational AI comprises a modest 16 members. Nevertheless, the company is allocating substantial resources, amounting to millions of dollars per day, toward training its language models. Training these extensive language models necessitates a significant investment in hardware infrastructure. For context, OpenAI’s Sam Altman disclosed that the development of GPT-4 incurred expenses exceeding $100 million.
Beyond chatbots, Apple has laid out additional objectives in the AI domain. One of these ambitions involves the creation of a feature enabling voice assistants like Siri to automate complex, multi-step tasks. Although rudimentary versions of this capability exist on iPhones today, they currently require manual configuration through the Shortcuts app. There’s speculation that the Siri team may introduce voice-controlled automation for multi-step tasks as part of iOS 18.
Furthermore, Apple has dedicated AI teams working on software to generate videos, images, and multimodal AI that seamlessly integrates text, images, and videos. Notably, the Ajax chatbot Apple is developing purportedly surpasses the capabilities of the original ChatGPT 3.5, boasting training on an astounding 200 billion parameters. However, OpenAI’s newer models have since eclipsed this achievement, showcasing the ever-evolving landscape of AI technology.
Conclusion:
Apple’s aggressive investment in conversational AI showcases its determination to remain competitive in a rapidly evolving market. The company’s pursuit of automation and multimodal capabilities signals its readiness to meet the growing demand for advanced AI solutions. As the AI landscape continues to advance, Apple’s strategic positioning positions it as a formidable player in the industry, ready to capitalize on emerging opportunities.
