The Limitations of the OpenAI Charter
According to Bard, ChatGPT, and Claude
The OpenAI Board Drama is a thanksgiving gift for which we should all be grateful. The tensions between nonprofit safety goals and commercial motives propelling AI research and product development along with the board and management conflicts mirror what Adam Smith described as the human "propensity to truck, barter, and exchange." Money, health, and human drama weave together to demystify artificial intelligence and reveal a story to which many more people can relate.
Our Charter describes the principles we use to execute on OpenAI’s mission
Bard, ChatGPT, Claude, three generative AI tools helped me assess the OpenAI Charter. The charter plays a crucial role in board governance. The Board is responsible for ensuring that OpenAI's work is safe, transparent, and accountable. The pre-Thanksgiving drama at OpenAI resulted in a restructuring of the Board. Three former board members, Helen Toner, Ilya Sutskever, and Tasha McCauley are gone. The new board is made up of former Treasury secretary Larry Summers, former Salesforce co-CEO Bret Taylor as chair, a nonvoting seat for Microsoft, and the only former board member to remain is Quora CEO Adam D'Angelo. I share the limitations in the OpenAI Charter. Risk management is especially noteworthy and I offer a solution to the question: How should the new board manage the risk of OpenAI products such as ChatGPT and Dall-E?