Elon Musk has a vision for Tesla that goes beyond just making cars. He wants Tesla to be a leader in artificial intelligence (AI), specifically in developing technology that allows cars to drive themselves without human intervention.
At the heart of this ambition is a powerful computer called Dojo. This supercomputer is specially designed to help train Tesla’s Full Self-Driving (FSD) systems. It’s important to note that while FSD can handle some driving tasks automatically, it still needs a human driver to be alert and ready to take control. Musk believes that with more data, better computing power, and extensive training, Tesla can evolve from having cars that are almost self-driving to ones that can drive themselves completely.
Musk has been talking about Dojo for a while now, and in 2024, he started discussing it more frequently. As we move into 2025, another supercomputer called Cortex has also been introduced, but Dojo remains crucial for Tesla, especially as electric vehicle (EV) sales have slowed down. Investors are looking for reassurance that Tesla can achieve fully autonomous driving.
Here’s a timeline of key moments in the development and discussion of Dojo:
2019: The Beginning
- April 22: At an event called Tesla’s Autonomy Day, Musk first mentions Dojo. He describes it as a supercomputer that will help train AI for self-driving cars. He also mentions that all Tesla cars produced at that time had the necessary hardware for full self-driving, needing only a software update to activate it.
2020: Building Momentum
- February 2: Musk announces that Tesla will soon have over a million connected cars worldwide, equipped with sensors and computing power for full self-driving. He highlights Dojo’s capabilities, stating it will process vast amounts of video data.
- August 14: Musk calls Dojo a “beast” and predicts it will be ready within a year, aiming for an August 2021 launch.
- December 31: Musk emphasizes that while Dojo isn’t strictly necessary, it will improve the safety and performance of self-driving technology.
2021: Official Announcement
- August 19: Tesla officially announces Dojo at its first AI Day, aiming to attract engineers. They introduce the D1 chip, which will be used in the Dojo supercomputer, and mention plans for a large AI cluster with 3,000 D1 chips.
- October 12: Tesla releases a technical document explaining how Dojo will work with deep learning neural networks.
2022: Progress Updates
- August 12: Musk mentions that Tesla will gradually integrate Dojo into its operations, reducing reliance on other types of computing hardware.
- September 30: At the second AI Day, Tesla reveals it has begun testing Dojo and showcases its capabilities, including generating an AI image of a Cybertruck on Mars.
2023: A Risky Venture
- April 19: Musk describes Dojo as a high-risk but potentially rewarding project that could significantly reduce training costs for AI.
- June 21: Musk announces that Dojo is already operational and running tasks, with plans to scale its capabilities significantly.
- July 19: Tesla reports it has started producing Dojo and plans to invest over $1 billion into it through 2024.
2024: Scaling Up
- January 24: Musk reiterates that Dojo is a high-risk project and mentions plans for future versions, indicating ongoing development.
- July 23: Musk highlights challenges in obtaining Nvidia hardware, stating that Tesla needs to focus more on Dojo to ensure they have the necessary training capabilities.
2025: New Developments
- January 29: In Tesla’s earnings call for 2024, there is no mention of Dojo, but the new Cortex supercomputer is discussed. Cortex, made up of about 50,000 Nvidia GPUs, is expected to enhance Tesla’s self-driving technology.
In summary, Elon Musk’s goal is to transform Tesla into a major player in AI, particularly in creating fully autonomous vehicles. The Dojo supercomputer is a key part of this plan, designed to process vast amounts of data and improve Tesla’s self-driving capabilities. However, as new projects like Cortex emerge, the future of Dojo and its role in Tesla’s AI ambitions remains to be seen.
