From Vox Media: 2024-06-05 15:15:00

Employees from leading AI companies are requesting “a right to warn about advanced artificial intelligence” to prevent potential harm. The 13 signatories, including former OpenAI and Google DeepMind employees, are concerned that current safeguards aren’t enough to mitigate risks.

The new proposal seeks to change incentives for AI companies by allowing employees to voice concerns transparently without fear of retaliation. The signatories, some of whom are anonymous, want to address potential risks without disclosing intellectual property and trade secrets. They aim to enhance accountability and transparency in the industry.

AI safety researchers emphasize the need to realign incentives in the AI industry to prioritize safety over rapid development and profit. Former employees of OpenAI have raised concerns that profit motives are overshadowing safety protocols. The industry’s emphasis on competition and commercialization creates challenges in maintaining safety protocols.

Efforts to establish unique governance structures in AI companies, such as nonprofits with a mission to protect humanity’s best interests, have not entirely solved the problem. The costly nature of AI experiments has led many organizations to prioritize profits, compromising safety measures. Advocates stress the importance of structural changes to foster a safety-centric culture within the industry.



Read more at Vox Media: OpenAI insiders have signed an open letter demanding a “right to warn” the public about AI risks