Apple is working on an AI chatbot of its own

Apple's secret project hinges on a framework the company calls "Ajax", which will be used to create Large Language Models (LLM), similar to OpenAI's GPT models. (Representative Image)

Apple has been secretly working on AI tools to challenge Google, OpenAI and Microsoft. The Cupertino-technology giant has been quiet about its plans on how it intends to compete in the AI space but thanks to a new report, we finally know what the company has been working on.

Apple's secret project hinges on a framework the company calls "Ajax", which will be used to create Large Language Models (LLM), similar to OpenAI's GPT models.

According to Bloomberg, Apple is using the framework to create a chatbot that some engineers refer to as "Apple GPT", internally.

Employees have already begun using the chatbot for help with prototyping, and it can also summarize text or answer questions based on the data it has been trained with.

While other companies have gone all-in on AI, Apple has remained cautious. It has introduced some AI-based features like improving the keyboard with word and phrase prediction or how FaceTime can now use FaceID to generate effects and backgrounds during a video call.

In April, news broke that Apple was struggling internally with a direction for Siri, its virtual voice assistant on devices.

Apple has been trying to overhaul Siri since 2019 but the company's strict stance on privacy has posed a lot of challenges, like the requirement to keep and process data on-device, instead of the cloud like Google, OpenAI or Microsoft.

In the same month, another report said that Apple was testing an AI-based health coaching service that would use data from the Apple Watch to personalize training regiments for users.

Through it all, Apple CEO Tim Cook has been cautiously optimistic, saying that the technology has huge potential but "a number of issues need to be sorted".