recently,AmazonAnnounced a major strategic move to acquire a robotics startup Covariant and brought about a quarter of its employees under its wing.
The core of this transaction is that Amazon hopes to use Covariant's robotics foundation model to enhance its innovation capabilities in the field of automation. Covariant's founding team, including Pieter Abbeel, Peter Chen and Rocky Duan, will join Amazon's artificial intelligence and robotics team in Silicon Valley.
Note: The picture is from Covariant official video screenshot
Amazon currently has deployed as many as 750,000 robots in its logistics centers, and this acquisition aims to further improve the safety and adaptability of these robots. Amazon said that through cooperation with Covariant, it is expected to accelerate innovation in the field of automation and attract more top artificial intelligence talents.
Although Covariant will transfer some of its talent to Amazon, it will continue to support existing customers and continue to develop its warehouse automation technology. However, the overall departure of the founding team makes it feel like this acquisition is almost a complete acquisition, although the specific financial details have not yet been disclosed. It is reported that Covariant was valued at US$625 million in its last round of financing in April 2023.
Not only Amazon, but other tech giants such as Apple, OpenAI and Google DeepMind are also increasing their investments in robotics and artificial intelligence. For example, Apple is developing a desktop device with a robotic arm and a large display, which is expected to be launched in 2026 or 2027. The device is designed to simplify daily tasks, such as adjusting the screen angle during video calls or browsing recipes. OpenAI has partnered with the robotics company Figure to launch its latest humanoid robot Figure02, and plans to promote humanoid robots in the workplace and at home. Google DeepMind is constantly developing robots and their control models that can cope with complex environments, demonstrating the ability to navigate freely in new environments.
These technology companies' expectations of combining robots with large language models are to improve efficiency and open up new application scenarios, especially in industry and daily life. Whether this integration can work effectively in practical applications remains to be seen.