OpenAI is a nonprofit technology research lab based in San Francisco and has established itself as a leading voice in the artificial intelligence field. Their most notable achievement so far is the development of the advanced AI model ChatGPT and its impressive capabilities. The advancement of OpenAI has been credited to its Chief Technology Officer, Mira Murati, who joined the lab back in 2018.
Murati is the driving force behind OpenAI’s ambitious projects and leads their research, product and safety teams. Under her guidance, OpenAI launched DALL-E, their image generator, and the most recent innovation, GPT-4. Murati recently spoke with the Associated Press about OpenAI’s vision of Artificial General Intelligence (AGI) and safety measures they take to ensure AI models are in line with human intentions.
Murati sees AGI development as the ultimate goal of OpenAI and they are striving to create systems that can reach their full potential while maximizing the benefits to society. OpenAI’s approach to safety involve making sure their AI models understand a wide variety of data not just text. They went so far as to curate their training data, manually adjusting the ratio of male to female images they receive. They also apply reinforcement learning with human feedback to get the AI system in line with what humans prefer.
For Murati, strong regulation is a must for AI systems. As OpenAI works with governments and other players in the field, they are trying to develop standards for the deployment of models that are both safe and low risk. In addition, Murati does not believe that a flat-out ban on the development of AI models more powerful than GPT-4 is the answer. According to her, agreements are needed between the industry’s players to build safety mechanisms and coordinate the evolution of AI systems.
In the last few years, OpenAI’s mission and structure has changed drastically and Murati was quoted as saying that while their intrinsic motivation and mission aligned attitudes have not changed, ‘their thinking has evolved a lot when it comes to what’s the best way to deploy, what’s the safest way’. As Murati continues to foster OpenAI’s technological growth, so does their focus on safety, making sure to find a delicate balance between advancement and security.
