Businessman and investor Elon Musk expressed his opinion in a recent post on social media platform X, advocating for the passage of California’s Senate Bill 1047, or the Safe and Secure Innovation for Frontier Artificial Intelligence Models Act. Elon Musk, who has been in favor of AI regulation for more than two decades, emphasized that such regulation is necessary for any technology or offering that poses possible risks to the public.
In response, Ethereum blockchain co-founder Vitalik Buterin expressed concerns about how the bill might impact the open-source weight model and requested more substantial evidence. He mentioned that earlier versions of the bill included a requirement for a complete shutdown, which conflicted with the open-source weight model, although this requirement has since been removed. Vitalik Buterin also noted that some AI safety advocates have supported the idea of banning open-weight models. He further suggested that a charitable interpretation of the bill’s objective could be to mandate safety testing, and if testing reveals potentially harmful capabilities or behaviors, those models would be prohibited from release.
Senate Bill 1047 seeks to create a regulatory framework for managing large-scale AI models. The bill targets models that exceed specific compute and cost thresholds. It holds developers legally responsible for the subsequent use or modification of their models. Before training begins, they must certify that their models will not provide or enable “hazardous capabilities” and must implement various safeguards to prevent such outcomes. Enforcement of these regulations would be managed by a newly established “frontier model division,” responsible for setting safety standards and advising on AI-related legislation.
What's the best evidence that the bill is going to be used to go after open weights?
I know that earlier versions of the bill had a full shutdown req that's incompatible with open weights, but that's been removed. And I know that some AI safety people have expressed support for…
— vitalik.eth (@VitalikButerin) August 27, 2024
SB 1047 Faces Backlash From AI Industry, California Assembly Vote Approaches
The introduction of SB 1047 has sparked strong reactions from the technology industry, with many prominent AI companies and organizations voicing concerns about its potential effects on innovation and their operations. Major entities such as AI research organization OpenAI and AI safety and research company Anthropic, along with several members of California’s Congressional delegation, have publicly opposed the document. They argue that AI regulation should be managed at the federal level rather than through a patchwork of state-specific policies. In response to this industry feedback, the bill has been revised several times to address these concerns while preserving its fundamental goals.
The California Senate passed Bill 1047 in May, which is now scheduled for a vote in the California Assembly in August. Should it pass, it would require only the signature of Governor Gavin Newsom to become law in California.
The post Elon Musk Backs California’s AI Safety Bill, But Vitalik Buterin Warns Of Risks To Open Source Models appeared first on Metaverse Post.
