The Future of AI: AI vs. AI-Inspired Threats
Date: 2024-09-30 Author: Gabriel Deangelo Categories: IN WORLD
Nvidia CEO Jensen Huang said that new technologies based on AI are needed to combat the threats posed by artificial intelligence.
“Fighting the negative aspects of AI will require using AI itself,” Huang said at an event hosted by the Bipartisan Policy Center on Sept. 27 in Washington.
He noted that because AI can generate false information and data at incredible speed, combating these phenomena also requires similar speed and accuracy. Huang drew parallels with cybersecurity, saying that virtually every company is under constant threat of being hacked, and only the development of technologies, including AI, can help protect data and prevent attacks.
The Big Threat of AI and the US Election
The issues surrounding AI and its impact on disinformation are becoming increasingly pressing ahead of the US election in November. According to a Pew Research Center survey, nearly 60% of Americans, regardless of their political affiliation, are concerned that AI could be used to create false information about presidential candidates. At the same time, about 40% of respondents believe that AI will be used for negative purposes.
Recently, it was revealed that Russia and Iran are using AI to alter videos of Kamala Harris’ speeches, raising even more concerns about interference in the election process. Huang emphasized the importance of the US government not only regulating AI, but also actively using it in its work. He pointed to the importance of implementing AI in key departments such as the Department of Energy and the Department of Defense, and proposed the idea of creating an AI-powered supercomputer.
Future Growth of AI Energy Consumption
Huang also noted that future AI development will inevitably lead to a significant increase in energy consumption. He suggests that in the coming years, data centers where AI models are trained will consume 10 to 20 times more energy than today. According to the International Energy Agency, data centers already consume up to 1.5% of the world's electricity.
The head of Nvidia also suggested that in the future, AI will train other AI models, which will further increase the load on energy resources. One solution could be to create data centers in areas with excess energy, which will allow for the efficient use of local resources.
Thus, Huang sees the future of AI development as an integral part of the fight against its negative manifestations, emphasizing the need to use advanced technologies to protect and develop key infrastructures.
