Jensen Huang, CEO of Nvidia, has cautioned that China already possesses the processing power and data centre capacity required to train an AI model comparable to Anthropic’s Claude Mythos, which might pose a danger to international cybersecurity.
In an interview on the Dwarkesh Patel podcast on Wednesday, Huang was questioned about whether the US national security may be threatened by the Chinese government’s access to chips to train a model like Claude Mythos, which has cyberoffensive capabilities.
According to Huang, Mythos was trained in a “fairly mundane capacity”.
“You just need to acknowledge that chips exist in China to understand the amount of capacity and the type of compute it was trained on.”
After discovering thousands of software flaws in popular operating systems and browsers, Anthropic restricted access to their new AI model in April, raising concerns about its abuse in cyberattacks. If utilised improperly, a Chinese-made AI model with the same capabilities may cause chaos.
China possesses “enormous” amounts of computing power, according to Huang.
“They have fully powered, completely empty data centres.” They have ghost cities and phantom data centres, you know. Their infrastructure capacity is enormous. They could just gather more chips if they so desired.
Huang went on to say that China has an abundance of energy, some of the top computer scientists, 50% of the world’s AI researchers, and produces 60% of the world’s mainstream chips.
Source: X.com
Stay informed with the latest trends in Web3, blockchain innovation, and cybersecurity updates at 3verseTV
You need to login in order to Like
Leave a comment