Top News

The debate in the tech industry has intensified after Anthropic CEO Dario Amodei’s appeal to slow down Frontier AI development. Sam Altman, Elon Musk and Demis Hassabis supported the security measures, while some leaders raised questions.

silicon Valley: Anthropic CEO Dario Amodei’s appeal to control the pace of development of frontier AI models has sparked a new debate in the tech industry. OpenAI CEO Sam Altman, SpaceX chief Elon Musk and Google DeepMind’s Demis Hassabis supported security and the need for independent evaluation, while David Sacks and Cohere CEO Aidan Gomez questioned the intent and competitiveness of the proposal.

Dario Amodei expressed concern over the speed of AI

Anthropic CEO Dario Amodei published an essay titled ‘We Must Pace the Frontier’ on September 12. In it he argued that AI is advancing so rapidly that it is becoming difficult for researchers to develop it safely.

Amodei noted two major warning signs. These include signs of AI’s early abilities to improve itself and the recent OpenAI-Hugging Face incident. He also put forward a three-step plan including a system for independent evaluation of Frontier AI models.

They recommend that third-party evaluators be given almost the same access to Frontier AI models as employees, so that they can independently assess their capabilities and potential risks.

Sam Altman and Elon Musk supported

OpenAI CEO Sam Altman supported Amodei’s proposal, saying that the pace of development of Frontier AI needs to be controlled. He also praised the idea of ​​giving independent evaluators the same access as employees and said OpenAI would do so.

Altman said on September 13 that he wants a federal framework that has uniform safety standards for frontier AI. However, he also clarified that it is not necessary to wait for any anti-trust exemption or law to start this work.

SpaceX chief Elon Musk also supported Amodei’s proposed voluntary slowdown. Musk said, “Dario is right.” Even in 2023, Musk was among the prominent people to appeal to stop AI development.

Satya Nadella and Demis Hassabis also supported the security measures

Microsoft CEO Satya Nadella supported independent auditors and broader efforts to AI security. However, he said that there is a need to rapidly spread the benefits of AI to different countries, communities and companies.

Nadella said that both closed and open-source models should flourish in the Frontier AI ecosystem. They also announced a Code of Conduct for Microsoft’s in-house MAI models.

Google DeepMind chair Demis Hassabis praised the direction of Amodei’s essay. He said that the proposal to create an industry-wide standards body for Frontier AI has also been put forward for this reason. According to him, this institution can work on preparing evaluation protocols and safety tests with relevant federal agencies and US National Labs.

David Sacks raised questions on the intention of the proposal

US President Donald Trump’s technology adviser David Sacks questioned Amodei’s proposal. He said that if AI companies want to slow down the development then why do they need government intervention for this.

Sacks also alleged that it is not correct to characterize the demand for a slowdown as purely altruistic. According to him, if superintelligence is not to be developed then companies can agree to do so themselves. Making the preferred regulatory framework a condition is a different matter.

Hugging Face and Cohere also expressed different concerns

Clément Delangue, CEO of Hugging Face, said that AI safety cannot be solved in the closed rooms of a few select frontier AI companies. Hugging Face was among the affected platforms in an incident involving misaligned agents of OpenAI. The company has also launched the Open Alignment Initiative and has expressed interest in joining Anthropic’s embedded evaluators program.

At the same time, Cohere CEO Aidan Gomez raised the question whether some influential AI companies of Silicon Valley should set the rules and safety standards of this technology for the whole world. He said that such companies are demanding changes in competition rules and setting their own standards.

Jensen Huang has not yet responded to Amodei’s proposal.

Nvidia CEO Jensen Huang has not yet responded to Amodei’s current appeal. However, he has previously taken a critical stance regarding the security concerns of AI companies.

Huang recently said that AI companies are giving priority to security concerns to increase demand for cybersecurity products. He has also previously criticized Amodei’s alleged stance that only he has the ability to develop AI safely. At that time, Amodei had completely rejected this allegation, calling it a serious allegation.

It is clear from these different views that have emerged regarding the speed of Frontier AI that differences of opinion still persist in the industry regarding the balance between security, government regulation and competition.