OpenAI CEO urges regulation of big language models, warns of generative AI risks

Sam Altman, CEO of Microsoft-backed OpenAI, has called for U.S. regulation of the deployment of advanced large-scale language models, warning of the dangers of generative artificial intelligence without a solid policy framework.

During a hearing with the U.S. Senate discussing the governance of generative AI, Altman argued that as concerns about threats to society grow, regulation is critical to mitigating the risks posed by increasingly powerful models.

The threats discussed in his testimony centered on the spread of misinformation and violations of data privacy laws related to how OpenAI trains its models.

Senator Richard Blumenthal said at the hearing that the prospect of insufficiently trained artificial intelligence is "terrifying" and argued that new technologies must be "held accountable."

He pointed out that OpenAI’s ChatGPT has the ability to imitate and simulate real human interactions, and he began his testimony with a piece of speech generated by a speech-trained model.

Altmann suggested that creating a government agency to oversee the training and deployment of artificial intelligence would aid development, explaining the need for a "combination of licensing and testing requirements" for developers.

He also proposed revoking the licenses of developers who develop AI tools that exceed certain "thresholds" or "capabilities," such as models that replicate themselves or produce harmful content.

Altman advocates developing regulatory rules around how machine learning tools can generate responses from Internet mobile data.

“Users should be able to opt out of having their data used by companies like ours or other social media companies,” he added.

C114 Communication Network Jiang Junmu

The above is the detailed content of OpenAI CEO urges regulation of big language models, warns of generative AI risks. For more information, please follow other related articles on the PHP Chinese website!