OpenAI is doing a lot to lower the risks that come with its technology as they grow. The group is actively searching for a new Head of Preparedness. The main job of this high-level position is to find and reduce potential risks from advanced AI systems.
This executive will be responsible for finding threats in several important areas. These include effects on mental health and the safety of computers. The fact that OpenAI is trying to hire people shows that it knows its models are starting to cause big problems.
Understanding the role at OpenAI
The new Head of Preparedness at OpenAI will have a very important job. This person will follow the company’s official plan for keeping an eye on big risks. They will learn about powerful tools that could hurt a lot of people if they are used wrong. According to the job ad, this job is part of the company’s official Preparedness Framework.
The pay for this important job is $555,000 plus stock in the company. This shows how much responsibility the job has. The leader will deal with risks that range from immediate threats like advanced cyberattacks to more speculative future threats. Experts at the Center for Security and Emerging Technology (CSET) have written about general AI security principles.
The OpenAI team that worked on this project was put together for the first time in 2023. But the leadership has changed. Aleksander Madry, the former leader, transitioned to a new role focusing on AI reasoning less than a year after the team’s establishment. Other safety managers have either left the company or moved to jobs outside of the safety teams.
In his announcement, CEO Sam Altman talked about specific problems. He discussed models that are very good at finding bugs in software. He also discussed how AI could affect mental health, which is a worry that is becoming more common. The American Psychological Association (APA) and other groups have information about digital well-being.
Not too long ago, OpenAI changed its rules about safety. One thing is very important in the new framework. If a competing lab puts out a high-risk model that doesn’t have the same safety features, the company might change its own safety rules. This means that market competition could lead to changes in safety rules.
In the end, OpenAI needs to find a new Head of Preparedness. It indicates that the company is serious about keeping the powerful technology it makes under control. The main goals of this work are to earn people’s trust and get ready for a future where AI will change everything and need a lot of attention. The leader’s performance will determine how well the group can keep its promise of safe and useful AI.
