OpenAI Loses Two Safety and Governance Researchers: Departures Shake the AI Community
Daniel Kokotajlo and William Saunders, former members of the OpenAI team responsible for ChatGPT, have parted ways with the company. Kokotajlo expressed concerns on a public forum regarding OpenAI's future conduct in the era of AGI, emphasizing a lack of confidence in the organization's ability to act responsibly. Notably, Kokotajlo served on the governance team, while Saunders contributed to the Superalignment team during their tenure at OpenAI.
.jpeg "OpenAI Loses Two Safety and Governance Researchers: Departures Shake the AI Community")
Two key figures from OpenAI, focusing on safety and governance within the realm of ChatGPT, recently bid farewell to the organization.
.jpeg "OpenAI Loses Two Safety and Governance Researchers: Departures Shake the AI Community")
Daniel Kokotajlo departed last month, while William Saunders made his exit in February, as confirmed by sources familiar with the situation, who preferred to remain anonymous but are known to Business Insider.
Kokotajlo, an integral member of the governance team, notably served as an adversarial tester for GPT-4, launched in March of the previous year. Meanwhile, Saunders contributed to the Alignment team, later rebranded as the Superalignment team, since 2021.
In a candid disclosure on the online forum LessWrong, Kokotajlo articulated his decision to resign, citing a loss of confidence in OpenAI's ability to act responsibly as AGI approaches. He also shared insights into the discourse surrounding a potential pause in AGI development, expressing disillusionment with the prevailing sentiments.
Saunders, reflecting on his tenure, noted his resignation after three years with the ChatGPT developer. His responsibilities extended to managing the interpretability team, dedicated to ensuring AGI's safety and understanding the intricacies of model behavior.
These departures coincide with a series of exits from OpenAI, including executives Diane Yoon and Chris Clark, as reported by The Information. Additionally, researchers Leopold Aschenbrenner and Pavel Izmailov also bid adieu to the organization, signaling a period of transition within the AI research community.
Tags:
AI
.jpeg&description=OpenAI Loses Two Safety and Governance Researchers: Departures Shake the AI Community){kind=link}