However, Apple is unlikely to use OpenAI models for its AI features in iOS 18, instead testing its own AI models against ChatGPT. The SiriSummarization framework can do summarization using on-device models, with Apple using its own AI models for this framework and comparing its results against ChatGPT. The iOS 17.4 code suggests Apple is testing four different AI models, including an internal model called “Ajax.” The company is ramping up efforts to integrate large language models into iOS, developing its own system and comparing the results to models like ChatGPT and FLAN-T5.
Key takeaways:
- Apple is expected to introduce new artificial intelligence features with iOS 18, and is currently testing a new version of Siri powered by large language model technology.
- Apple is using OpenAI’s ChatGPT API for internal testing to help the development of its own AI models, as seen in the first beta of iOS 17.4.
- iOS 17.4 includes a new SiriSummarization private framework that makes calls to the OpenAI’s ChatGPT API, which Apple is using for internal testing of its new AI features.
- Apple is testing four different AI models, including its own internal model called “Ajax,” and is comparing the results of its system to things like ChatGPT and FLAN-T5.