© ElectrekCo / X / Fair Use The Tesla company made a rather surprising decision to shut down the Dojo project - an advanced supercomputer that has been developed for years to support autonomous driving and the development of humanoid robots. This change in strategy marks the end of the ambitious idea of securing a technological advantage in the field of artificial intelligence with one's own chips. ADVERTISEMENT From high expectations to quiet redistribution: Dojo was often presented as a key asset of the company In public, Dojo was often presented as a key asset of the company, based on a special D1 chip. Its purpose was to process massive amounts of video data that cars generate every day, all with the aim of improving Autopilot and the advanced Full Self-Driving system, as well as the operation of the robot named Optimus. Analysts then believed that this kind of technology could significantly increase the company's market value. In a 2023 report, Morgan Stanley estimated that “Dojo could add up to $500 billion to Tesla's value.” However, things seem to have changed. The engineers who worked on Dojo will now be reassigned to other data center projects, while the focus shifts to collaboration with external partners such as Nvidia, AMD and Samsung. ADVERTISEMENT According to available information, the contract with Samsung is worth about $16.5 billion and includes production until 2033 at a new facility in Texas. The decision to shut down the Dojo program comes at a time when the company is going through a turbulent period – declining sales, increasing competition, and the departure of key personnel such as Peter Bannon and David Lau. Some former team members have already moved to the startup DensityAI, where they continue to work in the field of AI infrastructure. Elon Musk himself stated in early 2024: “Dojo represents a long-term risk with an uncertain outcome, although with a potentially very high return.” It seems that they have turned to other visions and ideas.