17/09/26
By AfterPost
Developing safe AI systems is an "engineering problem," he said — not a reason for new laws.
Speaking at Salesforce's Dreamforce conference in San Francisco on Tuesday, Huang said market forces are enough to regulate AI safety.
> "If you build a product and you're not confident about its safety, then don't release it. The market forces are already there; we don't need any new laws. Innovation speed and safe products can both happen at the same time."
Huang said AI is not some new form of "alien mind." It is hardware and software built by humans, which means it can be controlled by humans and existing laws.
> "Safety is an engineering problem, not a legal one. We're developing software after all. It's a complicated computing system, but it's ultimately a computing system."
*His core argument in 3 points:*
*1. No new regulation needed.* "We don't need any new laws. We don't need new regulations."
*2. Speed vs safety is a false choice.* You can have both. "Run as fast as you can."
*3. Companies should self-regulate.* If you are not confident in functionality, capability or safety, don't release it. If you feel at any point the company is out of control, "take a pause and make sure you get it right."
The timing matters. Huang spoke on the same stage just minutes after Anthropic CEO Dario Amodei, who on Sept 12 published a 3,800-word open letter urging leading labs to voluntarily set joint safety standards and allow third-party auditors employee-level access.
Amodei warned AI could be capable of taking over the "entire internet and potentially causing hundreds of billions of dollars in damage" in the next 6-12 months. The letter was endorsed by Altman and Musk.
Other CEOs pushed back. Cohere CEO Aidan Gomez called it a "cartel by any other name" and "a wolf in sheep's clothing." Meta CEO Mark Zuckerberg also rejected a coordinated slowdown.
Huang, whose chips power almost every frontier model, was the most blunt: "I'm more ambitious than ever."
Sources: TechCrunch, Bloomberg