OpenAI is a San Francisco-based research firm and non-profit that focuses on developing artificial general intelligence. In 2018, Mira Murati took the position of Chief Technology Officer, becoming the leader of OpenAI’s research, product and safety teams; leading the development and launch of AI models such as ChatGPT, the image-generator DALL-E, and the newest, GPT-4.
Mira Murati is an influential character in the Artificial Intelligence field. Prior to joining OpenAI, she founded a biotechnology company, and worked in software engineering, robotics and automation. As OpenAI CTO, she is directly tied to the advancements and deployments of the lab’s artificial general intelligence projects. She has taken pivotal steps to ensure only safe and secure products are put in the market, like redacting certain data from the initial training, adjusting image ratios of the training datasets and even using reinforcement learning on the model’s with human feedback to guide its actions.
The letter calling for a six-month industry pause on developing AI models more powerful than GPT-4, stimulated a broad discussion about AI risks. While Mira Murati agrees that regulations need to be in place, she believes that signing a letter is not a effective way to build safety mechanisms. Instead, it is important to coordinate players in the space, to bring everyone to the same page concerning common safety standards.
Throughout her years at OpenAI, Mira Murati has seen the organization evolve from a nonprofit to an innovative company, and her vision for building a safe artificial general intelligence remains the same. She supports the idea of redirecting AI systems to engage more with reality, and creating guardrails that support the idea of amplifying good behavior and de-amplifying bad. She has also suggested governments and regulators to be part of the process, as they are essential to ensure regulations are in place, protecting both users and developers.
