Recently, the United States' Google launched a prototype satellite into space. This satellite is outfitted with Tensor Processing Unit (TPU) chips. The primary objective of this launch is to evaluate how these chips perform in the space environment, thereby examining the practicality of establishing artificial intelligence (AI) infrastructure in space in the foreseeable future. In the realm of space technology, sending prototypes into orbit for testing is a common practice. It allows scientists and engineers to gather real-world data and assess the functionality of new technologies under the unique conditions of space. The Google team has successfully established communication with the satellite, and as of now, it is functioning as expected. Google has clarified that this launch represents just one phase of a long-term research initiative. The overarching goal is to delve into the possibilities of setting up scalable machine learning infrastructure in space, which could revolutionize how we utilize AI beyond Earth's atmosphere.