Article

Jensen Huang draws AI a safety red line: If customers develop unsafe software, NVIDIA may stop supplying chips

Dongcha Beating AI Flash, NVIDIA CEO Jensen Huang recently said in an interview with Bloomberg in London that if customers develop unsafe AI software, NVIDIA could hypothetically stop supplying them with chips, but would communicate with customers before taking action. Huang also emphasized that NVIDIA provides general-purpose technology and has certain limitations in judging how customers use it.

Meanwhile, NVIDIA has committed to provide the United Kingdom with computing resources worth 5 over the next 250000000 USD years to support AI safety research, with the relevant resources worth approximately 50000000 USD annually. Supported projects include the University of Oxford-led BOLD project, as well as related research at Imperial College London and University College London.

Huang advocated accelerating the integration of safety constraints and monitoring technologies into AI systems, stating that AI capabilities and safety must advance together. The report noted that AI safety regulation could affect demand for NVIDIA chips: if regulation restricts training of new models or raises development costs, it could curb purchases of AI processors; however, strengthened safety research could also help deploy AI technology more reliably.

Original link https://m.theblockbeats.info/flash/372350