Elon Musk has confirmed that cars with Tesla AI4 (also referred to as Hardware 4 or HW4) can achieve unsupervised self-driving without upgrading the hardware at all. The statement supports a long-standing strategy of Tesla that it already has a platform of vehicles that are capable of full autonomy because they do not need additional physical features, but instead, improvements in software and AI.
This was confirmed when a social media platform X user by the name @iliketeslas posted, stating that current AI4 hardware is already sufficient to run Unsupervised Full Self-Driving (FSD). Any person who has used AI4 currently can tell that unsupervised driving does not require superior hardware. It is so good at the moment, the post said. Musk replied in a very matter-of-fact way: Yup.
This recommendation is an indication of firm belief by Tesla leadership in the AI4 chip, which is already in use in the new vehicles of the company.
Musk’s Vision: AI4 to Surpass Human Safety Levels
Musk has already mentioned that self-driving safety will be reached with AI4 alone and will be much higher than human. This claim is consistent with the overall strategy of Tesla to rely on large-scale neural networks that have been trained with real-world data of driving to perform better than human drivers in safety and reliability measures.
AI4 by itself will achieve self-driving safety levels very far above human.
AI5 will make the cars almost perfect and greatly enhance Optimus.
AI6 will be for Optimus and data centers.
AI7/Dojo3 will be space-based AI compute.
— Elon Musk (@elonmusk) January 18, 2026
In contrast to other competitors, which depend on heavy use of lidar or regular hardware refresh cycles, Tesla still focuses on vision-based autonomy with more and more powerful onboard computers. Musk writes about AI4 as a huge inflection point in which the hardware ceiling is no longer the main bottleneck to autonomy.
Roadmap Beyond AI4: AI5, AI6, and AI7
Although AI4 is likely to open the unsupervised self-driving, Musk provided a roadmap for the future of Tesla chips. Musk states that the next chip, AI5, will transform Tesla cars into near perfection when it comes to autonomy and greatly improve the features of the Tesla Optimus humanoid.
In addition to that, AI6 will probably be targeting Optimus and massive data center applications, which will increase Tesla’s presence outside of the automotive applications. Musk claims that the most ambitious is AI7, also known as Dojo3, which will have the capability of doing space-based AI computation.
Space-Based AI and the Dojo Revival
The remarks of Musk also revitalize the Dojo project of Tesla, which is the company’s own AI training center that the company is working on. Despite being supposedly scrapped last year, Dojo was recently announced by Musk as Tesla reviving it in the shape of Dojo3, indicating a potential new dedication to in-house AI infrastructure.
Musk is not the first to think of space-based AI computers. He has already proposed that SpaceX would install data centers in orbit, which might be less costly and energy-efficient than those on the ground. When implemented in practice, this type of infrastructure would potentially transform the training and deployment of large AI models radically.
Besides, all these advancements of Tesla have raised concerns among other industry players. Gary Black, the managing partner at Future Fund LLC, warned that automakers who do not invest heavily in autonomous driving risk a BlackBerry moment. The source points out the speed with which BlackBerry became irrelevant as the smartphone market shifted to touchscreen phones in the wake of the introduction of the Apple iPhone.
The point is obvious: autonomy can be one of the characteristics of future vehicles, and those companies that are behind AI and self-driving technologies can face the risk of going out of fashion quickly.
