From Verizon Media: 2024-05-18 12:01:00

OpenAI’s Superalignment team, tasked with governing superintelligent AI, was promised resources but faced denials hindering their work, leading to key resignations. Co-lead Jan Leike criticized the company’s priorities and lack of readiness for next-gen AI challenges. OpenAI formed the team to address crucial safety and alignment issues for AI development.

Despite OpenAI’s efforts, several key team members, including co-lead Jan Leike and co-founder Ilya Sutskever, resigned due to disagreements over priorities and leadership issues within the company. The Superalignment team aimed to solve essential challenges in controlling superintelligent AI, but faced obstacles in securing necessary resources and investment.

OpenAI’s struggle to balance product launches with safety-focused AI development led to internal conflicts and distractions. Co-founder Ilya Sutskever’s departure and subsequent disputes with CEO Sam Altman highlighted internal tensions. The company faced challenges in maintaining a safety-first approach in AI development amidst leadership and organizational changes.

Amidst the departures of key team members, OpenAI is reorganizing its approach to AI safety and alignment work. The Superalignment team will now be integrated into various divisions within the company, raising concerns about maintaining a strong focus on safety in AI development. The changes aim to continue the company’s mission amidst internal restructuring.



Read more at Verizon Media: OpenAI created a team to control ‘superintelligent’ AI — then let it wither, source says