Smartphone manufacturers are aiming to kickstart a new "super cycle" in the smartphone industry with the help of AI. They're betting on the future of smartphones that look less like traditional ones and more compact, enabling interaction via voice activation.
Google is optimistic about the prospects of employing AI large models in smartphones. Brian Rakowski, Vice President of Product Management at Google's Pixel division, recently projected that Google's more advanced Gemini large model will be embedded into smartphones next year.
Large language models are artificial intelligence models capable of understanding and generating language in a manner close to human-like. Google's top-tier large language model, Gemini Ultra, boasts a staggering 1.56 trillion parameters, while OpenAI's GPT-4 has 1.76 trillion parameters.
Google's Gemini is competing with OpenAI's GPT-4 model. Google has made progress in compressing models to fit mobile devices, currently offering a scaled-down version of Gemini called Gemini Nano on its Pixel devices and select Android devices. The higher-tier versions of Gemini are currently accessible only via the cloud, but Rakowski anticipates these versions will be available on Android phones starting in 2025.
"Gemini Nano's performance matches what our online models offered less than a year ago," Rakowski added, highlighting the myriad tasks people can accomplish with these scaled-down models. If this trajectory continues, smartphone users will be able to access and utilize AI large models more directly and rapidly. "It's instantaneous, without relying on cloud connections or subscriptions."
According to CNBC, global smartphone sales plummeted to 1.16 billion units in 2023, hitting a decade-low. Following years of significant slowdown in device sales, smartphone manufacturers are hopeful that AI will drive a new "super cycle" in the industry.
However, analysts suggest that a super cycle in the coming years is unlikely due to the lack of sufficient new features and innovations to persuade people to upgrade their smartphones. "We don't expect to see this kind of prosperity," said Francisco Jeronimo, Vice President of Data and Analysis at research firm IDC.
Nevertheless, an increasing number of smartphone manufacturers are investing in artificial intelligence, banking on the future of smartphones that look less like traditional ones and more compact, enabling interaction via voice activation. Google, too, is making substantial bets in the AI field, aiming to gain an edge in competition with rivals like OpenAI. Google recently rebranded its chatbot Bard as Gemini, with Android users able to download a dedicated Android app to use Gemini, while iPhone users can access Gemini via Google's iOS app.