Apple's focus on local AI models and licensing of LLM programs could be a winning combination, the analyst said.

Apple's AI push appears to be going both ways

In the analyst's note it says licensing Gemini from Google while developing its own private local models could give Apple an edge in the AI ​​race.

It's no secret that Apple plans to make 2024 a big year for promoting artificial intelligence technologies on its platforms. Recent reports have begun to give shape to Apple's artificial intelligence plans, and they are similar to other strategies the company has pursued in the past.

Apple's potential partnership with an LLM supplier to develop smaller local models could help the company gain a leg up on other manufacturers, according to a JP Morgan note reviewed by AppleInsider. If the reports are true, Apple will focus on its strengths, such as on-device data processing, instead of creating its own LLM.

A recent report revealed that Apple is in talks with Google to license Gemini LLM for iPhone. The report comes just a day after Apple published a research paper on MM1, a small pre-trained model that can run locally on a user's iPhone.

The news is complementary, not controversial, and JP Morgan's note said Apple's two-pronged approach could give it an advantage. Apple could focus on smaller device models that protect user privacy rather than releasing a controversial LLM program based on reams of data found online. Clients get the best of both.

If this sounds familiar, it's a lot like Apple's search deal with Google. Google becomes the default search engine on the Internet, and Apple provides a powerful but private local search tool called Spotlight.

Apple's plans are supported by further evidence: leaks about an internal tool called Ask, which is trained on a local knowledge database. It was said to be more adaptive than LLM and flexible when the database changes, which is what the MM1 model should be suitable for.

The note focuses on Apple's financial success using this approach, which involves infrastructure savings and improved app user experience. JP Morgan maintains an “outperform” rating on Apple with a price target of $215.

Leave a Reply

Your email address will not be published. Required fields are marked *