• TWEAK
• Computer
• Laptop

Apple updates Mac with AI via M4 chips

Translate from : Apple opdaterer Mac med AI via M4 chips

According to Bloomberg, Apple's artificial intelligence-centric M4 processors will be implemented in almost every Mac variant released by the iPhone maker. This is an attempt to increase sales of the Apple Mac line using AI features.

According to people familiar with the situation, the tech giant, which debuted its first Macs with M3 processors five months ago, is already close to starting production of the next version, the M4 CPU. According to sources who did not want to be named because plans have not yet been made public, Apple plans to replace existing Mac models with the new CPU, which will be available in at least three main variants.

Apple reportedly plans to launch the updated laptops in late 2024 or early 2025. The lineup will include new iMacs, Mac minis, high-end 14- and 16-inch MacBook Pros, as well as a low-cost 14-inch MacBook Pro . The new Macs are being developed at a crucial time. Mac sales fell 27% in the latest fiscal year that ended in September, after a boom in 2022.

Revenue from the computer line was flat during the holiday period. With an M3-focused launch event last October, Apple tried to revive the Mac industry, but these chips didn't much outperform the M2 from the previous year. The M4 chip family consists of a top-tier processor codenamed Hidra, more powerful Brava models and an entry-level model called Donan.

The upcoming macOS version, expected to be unveiled in June at Apple's annual developer conference, will showcase the company and its AI processing capabilities.

As the AI race continues, Apple continues to catch up, as seen by the announcement of its latest system, Reference Resolution As Language Modeling (ReALM), which it claims outperforms GPT-4 in certain areas.

According to the Apple team, the ReALM system is positioned as a superior product that outperforms current LLMs, especially in processing certain types of queries. This means that it is not simply an attempt to catch up with competitors.

The report from the Apple team claims that ReALM's special ability to decipher obscure references on the screen and retrieve conversational and background data makes it unique in offering more accurate answers to user queries.

ReALM seeks to improve the accuracy of its responses by better understanding the user's intent behind a query by leveraging contextual clues from the user's screen and ongoing device activities.

Extensive tests against several LLMs, especially GPT-4, according to the researchers, have shown that ReALM performs better on certain tasks.

Additionally, they suggest that Apple intends to integrate ReALM into its ecosystem, improving Siri's ability to offer more relevant answers. However, this may require an iOS 18 upgrade for consumers when it launches later this year.