When Mira Murati joined OpenAI in 2018, the San Francisco-based research lab was building a strong foundation of Artificial Intelligence (AI) while still far from becoming a household name. However after Murati joined the nonprofit, the lab made a major transformation and became popular worldwide for its AI models, especially ChatGPT.
Murati is the current CTO of OpenAI, leading the research, product and safety teams of the company. Not only has she been responsible for the development of AI models such as ChatGPT, DALL-E (an image-generator) and GPT-4 (the latest model), she also spoke to the Associated Press recently.
During her conversation with the Associated Press, Murati spoke about the concept of artificial general intelligence, also known as AGI. She mentioned OpenAI’s vision of having AGI that is safely built, aligned with human intentions, and is able make maximum benefit for everyone.
Murati also spoke about the safety measures undertaken by OpenAI when developing these models. They made sure to audit the datasets and remove any offensive material. For ChatGPT, they implemented reinforcement learning in order to make it more aligned with human preferences.
Murati believes that AI systems should be regulated and caution should be taken while developing them. Talking about the 6-month industry pause on making AI models more powerful than GPT-4, Murati disagreed with the idea of signing a letter for building the safety mechanisms. She said that the letter had some untrue statements about the development of GPT-4 and that there should be more coordination and exploration among the players in the space.
OpenAI has made a major transformation in the five years Murati has been with them and their thinking has also significantly evolved about how best to deploy these AI systems. They made sure to make their structure such that it serves their nonprofit mission.
Mira Murati is an AI and robotics expert with over 15 years of experience in mechanical engineering, product management, and software engineering. She has a graduate degree in Robotics and Automation from the Technical University of Catalunya, and a degree in Business Administration from La Salle Barcelona. Her current mission with OpenAI is to build safe AI systems that positively benefit society.
