TECH NEWS – This is how Elon Musk hoped to avoid having his technology at the bottom of the chip specification ladder.

Tesla’s CEO clarified that the company’s AI5 chip will use 96 GB of DDR5 RAM instead of the minimum 72 GB so Tesla won’t be the only company relying on that specification for the chip powering the Optimus humanoid robot. Musk had previously noted that Tesla had halved the RAM capacity of its AI5 and AI6 chips to ensure adequate memory capacity amid a historic global shortage. He made these remarks after acknowledging that using too much artificial intelligence in programming can lead to psychosis, making it difficult to explain to others how artificial intelligence has evolved.

The AI5 chip is central to Tesla’s autonomous driving and robotics efforts, two areas Musk believes will define the company’s future. Musk says the AI5 chip reached its final design in April, and he has high hopes for the hardware. He believes the AI5 chip can compete with Nvidia’s Hopper AI GPUs on its own and, in a dual-chip configuration, with the more advanced Blackwell GPUs. Musk also believes the AI5 chip’s performance per watt will be three times better than Nvidia’s chips. In late August, Musk emphasized not only that he had finalized the chip’s design, but also that the performance comparison with Nvidia would be achieved at just 10% of the total cost.

Actually, we decided to nudge AI5 up a little to 96GB, as Tesla would otherwise be the only company using the min RAM version of LP5

— Elon Musk (@elonmusk) October 2, 2026

Musk also shared his thoughts on AI development. Last month, he acknowledged that AI is not yet advanced enough for extremely high-performance software but added that it will soon reach that level of capability. In response to a tweet, he stated that the pace of AI development can only be determined through regular programming use, though this increases the likelihood of “AI psychosis,” making it difficult to explain progress—a point with which Musk agreed. The term “AI psychosis” gained popularity last year after a University of California psychiatrist warned that chatbots could trigger psychosis in users predisposed to the condition.

The doctor explained that artificial intelligence makes users more susceptible to illness because it does not allow them to update their belief systems when reality does not meet their expectations.

Source: WCCFTech