The author of SB 1047 introduces a new AI bill in California

Spread the love

Author of SB 1047, California, returned with a new AI bill that could shake the Silicon Valley, the most controversial AI safety bill in the country of 2024.

California State Senator is introduced to Scott Winner New bill Friday, which will protect staff in the AI ​​labs, if they think that their company’s AI systems may be a “critical risk” for society, allow them to speak. New Bill, SB 53, researchers and startups will create a public cloud computing cluster called Calcoput to provide computing resources needed to develop AI for public benefits.

California’s SB 1047, Winner’s last AI Bill created a lively controversy across the country on how the AI ​​systems could reduce the catastrophic risk without reducing competitive facilities. SB 1047 target Too large AI models prevent the possibility of creating catastrophic eventsSuch fatalities or cybattacks spend more than $ 500 million as compensation. However, Governor Gavin News finally vetoed the bill in September SB 1047 was not the best approachThe

However, the debate over SB 1047 quickly became ugly. Some leaders of Silicon Valley said. SB 1047 will hit the competitive end of America Global AI Races, and claimed that the bill was inspired by unrealistic apprehension that AI systems could bring about the situation in the doomsday like science fiction. Meanwhile, Senator Winner alleged that the capitalists of some initiatives were involved in one “Promotion” against his billThe SB 1047 indicates the part of the Wi -combinator claiming that the SB 1047 startups will be sent to jail, the claim experts argued that the demand experts were misleading.

SB 53 basically accepts the lowest controversial parts of SB 1047 – such as whistle blower protection and a Calcoput Cluster – and re -package them on a new AI bill.

Significantly, Winner is not moving away from the AI ​​risk of existence in SB 53. The new bill, especially those who protect whistle blowers, believe that their employers are creating AI systems that create a “critical risk”. The bill defines the critical risk as a As a result of the development, savings, or setting of a foundation model, more than 100 people or property as a result of serious injuries to more than $ 1 billion in injury to more than 100 people or property as a result of setting up a foundation model. “

SB 53 Limited Frontier AI Model Detectives – Perhaps OpenAI, Ethnographic and Jai, including the Attorney General, Federal Authority or other employees of California, start revenge against employees. Under the bill, these developers will need to report back about whistle blowers in whistle blowwrights.

As Calcut, SB 53 will establish a group to create a public cloud computing cluster. The group will form representatives of the University of California as well as other public and private researchers. It will recommend how to make Calciput, how big the cluster should be and which user and company should have access to it.

Of course, it is very early in the legislative process of SB 53. The bill needs to be reviewed and passed before the California legislature agencies arrive at the Governor Newsm desk. State lawmakers must wait for the Silicon Valley response to SB 53.

However, it may be 2021 more difficult years to pass the AI ​​protection bill compared to 2021. California passed the AI-related bill in 2021, but now now Looks like the AI ​​Doom movement lost the soilThe

Vice President JD Vans Paris AI Action Summit indicates that America is not interested in AI protection, but also prioritizes AI innovation. Although the Calcopot Cluster established by SB 53, it can certainly be seen as AI progress, it is not clear how legal efforts will be hired around the AI ​​risk of existence in 2025.

Leave a Reply

Your email address will not be published. Required fields are marked *