Former OpenAI executive Jan Leike has made headlines by joining Anthropic, OpenAI’s rival backed by Amazon. Leike’s decision to leave OpenAI comes just weeks after he publicly criticized the company for prioritizing “shiny products” over safety culture and processes.
Leike took to social media to express his enthusiasm for his new role at Anthropic and encouraged interested individuals to join him and his team. This move follows a string of departures from OpenAI, including policy researcher Gretchen Krueger, who shared similar concerns about the company’s direction.
In response to these departures, OpenAI, backed by Microsoft, has taken steps to address safety concerns. They have established a safety committee overseen by board members Bret Taylor, Sam Altman, Nicole Seligman, and Adam D’Angelo. Additionally, they have announced plans to test their “next frontier model.”
Leike, who co-led the Superalignment team at OpenAI, was responsible for ensuring the safety and controllability of superintelligent AI systems. However, following the resignations, the team was disbanded and its members were reassigned to core research roles.
Anthropic, founded in 2021 by former OpenAI employees and executives, has positioned itself as a safety-focused company compared to its competitors. With a $4 billion investment from Amazon, Anthropic utilizes Amazon Web Services’ cloud infrastructure for developing its large language models.
The departure of Jan Leike and other notable figures from OpenAI highlights the ongoing debate surrounding the prioritization of safety in AI development. As AI technology advances, it becomes increasingly crucial to ensure that safety measures are in place to prevent unintended consequences. The establishment of a safety committee at OpenAI and Anthropic’s emphasis on safety demonstrate a growing recognition of the importance of ethical AI development.
As the AI industry continues to evolve, it will be interesting to see how companies like OpenAI and Anthropic navigate the delicate balance between innovation and safety. The participation of experienced professionals like Jan Leike in these discussions will undoubtedly contribute to the ongoing development of responsible AI practices.
