Apple has been much more reticent than other tech companies about its artificial intelligence efforts. But a couple of papers published by researchers in recent weeks suggest they will be preparing new features for the iPhone. There are new tips rumors about a chatbot and a project about digital avatars.
The latest document reveals that Cupertino is developing a new way to run large language models (LLMs) on devices with limited DRAM memory. This would allow run these artificial intelligence systems directly on your iPhone or iPad, instead of doing it through external servers.
These language models that chatbots like ChatGPT run on typically run in data centers that have computing power that far exceeds that of an iPhone. But in a paper titled “LLM in a Flash,” the researchers say they have found “a solution to an existing computing problem.”
The proposed solution involves building a model that is in tune with the behavior of flash memory, focusing on two critical areas. On the one hand, reduce the amount of data transferred from flash memory. On the other hand, the report says that data is read in larger, continuous chunks.
Apple has tested its approach on models such as the Falcon 7B. It is a smaller version of a large open source language model originally developed by the Abu Dhabi Institute of Technology Innovation. “Such approaches will be important for harness the full potential of these language models across a wide range of devices and applications.”they note.
Apple is preparing its artificial intelligence for what will happen in 2024
Apple launched late last month MLX, a platform with which you can develop machine learning models.. According to rumors, this could become the basis for the development Generative application based on artificial intelligence. The system takes advantage of Apple’s own hardware so it doesn’t rely entirely on external processing.
Other reports last June said the company was working on its own chatbot, “Apple GPT,” as some engineers called it. The system will use a large language model (LLM) called “Ajax”. Apple was testing the tool internally, people familiar with the matter said. but without yet defining a clear strategy for bringing it to market.
Some competitors are already preparing to introduce artificial intelligence into mobile phones. Counterpoint calculated that More than 100 million AI-focused smartphones will be released in 2024. And that by 2027, 40% of new devices will offer this technology.
This month, Google unveiled a version of its new Gemini system that will run natively on Pixel smartphones. And in September, it was revealed that OpenAI, the company behind ChatGPT and DALL-E, wants to enter the consumer electronics market. The idea was discussed create an “iPhone of artificial intelligence”. It was even mentioned that Jony IveApple’s former head of design will join the project.
Digital avatars and what will happen in 2024
The second Apple study, published in late November, involves a new research project called HUGS. It is a generative artificial intelligence tool that can turn a short video of a person into a digital avatar in minutes. The HUGS method allows you to create animated digital avatars from just a few seconds of video. To be precise: from 50 to 100 frames.
These avatars can be animated and placed in different scenes as the platform uses separate representation of people and scenes. It’s easy to imagine how the iPhone’s camera and processing power could be used to provide this new personalization experience that could be applied to social media, gaming, or even augmented reality.
Qualcomm CEO Cristiano Amon has predicted that artificial intelligence will reverse the decline in mobile device sales. ““In early 2024, we will see the release of devices with a range of generative AI use cases,” said Financial Times.
Amon, the leader of the world’s largest mobile chip maker, believes the time will come to create more advanced virtual assistants that can predict more user actions. For example, send text messages or schedule a meeting. He also believes that there will be many new developments in the field of photo editing.
Source: Hiper Textual
I’m Ben Stock, a highly experienced and passionate journalist with a career in the news industry spanning more than 10 years. I specialize in writing content for websites, including researching and interviewing sources to produce engaging articles. My current role is as an author at Gadget Onus, where I mainly cover the mobile section.
