Compared to OpenAi or Google, Apple is behind in the field of artificial intelligence. While his house AI, Apple Intelligence is not as advanced as Chatgpt, Gemini or other similar systems, the company seeks to improve it and catch up. And, one of the ways to get there is to use customer devices, as revealed by a blog article published on Monday, April 14 and transmitted by Bloomberg.
In detail, the American giant will not use this information to train Apple Intelligence. Currently, their AI models are formed using synthetic data or information created to imitate real world data without any personal detail.
Consent
User data analysis will improve Apple Intelligence text functions, as notifications. Specifically, when synthetic emails are generated, an imprint will be created “that records the key dimensions of the message, such as language, matter and length” for each of the last.
These digital footprints will be sent to a small number of user devices that have been granted for this purpose, to compare it with real emails and determine which corresponds best to the emails of the real world.
Users can consent to the analysis of their data with the next beta versions of the Operating Systems of iPhone, IPAD and MAC.
Source: BFM TV
