Google’s DeepMind Quietly Hiring Nuclear Physicists

Futurism

Source published: Site published:

Original article · فارسی

An illustration shows researchers in protective gear observing a high-tech device.
Shutterstock / Design Projects

Leading AI labs have been hiring experts in many academic fields for years. They often use these experts as high-level fact-checkers to help train future AI models. But Google's latest job posting is unusual, even by these strange standards. It may also reveal a lot about what big tech companies have planned for humanity's future.

As the technology industry publication Data Center Dynamics reported, Google's AI research lab, DeepMind, is hiring a team of "highly qualified physicists and nuclear engineers." The job is for a "nuclear subject matter expert" on a team that deals with chemical, biological, radiological, and nuclear risks, as well as powerful explosives. This team is part of the company's Responsible Development and Innovation group.

According to the job posting, applicants should have a PhD in a field such as plasma physics or nuclear engineering, or have similar practical experience. They also need experience working with a US government laboratory and knowledge of policies related to managing nuclear weapons stockpiles, stopping the spread of nuclear weapons, controlling arms, or reducing threats. These are only the minimum requirements.

Google would also prefer applicants to have experience with AI safety. Other useful skills include computer-based scientific modeling, experience working at major laser research facilities, or the ability to design tests and safety measures that are clearly not classified and are not controlled by export laws. The job posting says this could involve using related scientific topics that have both peaceful and military uses as stand-ins for weapons-related topics.

Successful applicants could earn about $174,000 to $252,000 a year, plus a 15 percent bonus.

However, the job will not be easy. The posting says the work includes checking the safety risks of so-called "frontier AI models" when used in nuclear-related fields. More specifically, experts must examine whether these models could help people with nuclear fission, nuclear fusion, the spread of thermonuclear weapons, the movement of fluids affected by radiation, magnetohydrodynamics, and related subjects. The work must be done in ways that are clearly not classified and are not controlled by export laws.

Taken together, the details make it seem as though Google's leading AI lab expects its AI tools could begin producing practical instructions for developing nuclear devices. The company appears to be hiring nuclear physicists to help lead efforts to defend against these risks.

It is not yet clear how likely this situation is to happen. But Google is clearly prepared to spend money on addressing the possibility.

More on Google: The Tech Industry's Aggressive Push to Integrate AI Into Schools Is Turning Into a Disaster