Argmax has released a software package called WhisperKit, which allows OpenAI's Whisper speech recognition model to run smoothly on Apple Watch. This innovative integration is made possible by Apple's CoreML framework, enabling Whisper to showcase its capabilities on all Apple devices that support this framework. By leveraging the efficient performance of Apple's Neural Engine, WhisperKit can process voice data in real-time, bringing powerful speech recognition capabilities to WatchOS applications. It is worth mentioning that this software package is open-source under the MIT license and requires macOS 14.0 or higher and Xcode 15.0 or higher. Developers can easily integrate WhisperKit into their Xcode projects and flexibly choose audio formats and models according to their needs. One creative user even combined WhisperKit with his Vision Pro headphones to achieve transcription functionality. Since its establishment in 2020, Argmax has been deeply involved in natural language processing, recommendation systems, computer vision, and other fields. As an open-source project, WhisperKit encourages developers to actively contribute code to enhance its functionality and adaptability. The launch of WhisperKit by Argmax aims to expand the coverage of speech recognition technology in the Apple ecosystem and promote its widespread application in various types of applications. In recent years, the use of large language models on devices such as Apple Watch has become a trend, aiming to achieve on-device processing of complex tasks. This approach not only reduces latency and improves privacy protection but also provides users with a smoother interactive experience. At the same time, some unique devices such as Rabbit R1 and Humane Ai Pin have made significant achievements in integrating artificial intelligence. Rabbit R1 interacts directly with the user interface through a model, while Humane Ai Pin adopts an AI-based operating system that allows quick access to services without the need for wake words, with a focus on user privacy and convenience. These devices bring AI use cases to smaller and more efficient devices, making the technology more personalized and efficient.