OpenAI Wants to Pay Someone $555,000 a Year to Mitigate AI Downsides

OpenAI needs to pay somebody over half 1,000,000 {dollars} to mitigate the downsides of AI.

If that looks like some huge cash, take into account these potential downsides: job loss, misinformation, abuse by malicious actors, environmental destruction, and the erosion of human company, to call a number of.

CEO Sam Altman described the job as “irritating” in an X put up on Saturday. “You will soar into the deep finish just about instantly,” Altman wrote.

Altman mentioned the “head of preparedness” place is “a important function at an essential time.”

“Fashions are bettering rapidly and are actually able to many nice issues, however they’re additionally beginning to current some actual challenges. The potential influence of fashions on psychological well being was one thing we noticed a preview of in 2025; we’re simply now seeing fashions get so good at pc safety they’re starting to seek out important vulnerabilities,” he wrote.

OpenAI’s ChatGPT has helped popularize AI chatbots amongst shoppers, lots of whom use the expertise to analysis subjects, draft emails, plan journeys, or carry out different easy duties.

Some customers additionally speak to the bots as an alternative to therapy, which has exacerbated psychological well being points in some circumstances, encouraging delusions and different regarding habits.

OpenAI mentioned in October it was working with psychological well being professionals to enhance how ChatGPT interacts with customers who exhibit regarding habits, together with psychosis or self-harm.

OpenAI’s core mission is to develop synthetic intelligence in a manner that advantages all of humanity. It made security protocols a central a part of its operations from the outset. Because it started releasing merchandise, nevertheless, and stress to show a revenue grew, some former staffers have mentioned the corporate started to prioritize revenue over security.

Jan Leiki, the previous chief of the now-dissolved security group at OpenAI, mentioned in a Might 2024 put up on X asserting his resignation that the corporate had overpassed its mission to make sure the expertise is deployed safely.

“Constructing smarter-than-human machines is an inherently harmful endeavor. OpenAI is shouldering an infinite accountability on behalf of all of humanity,” he wrote. “However over the previous years, security tradition and processes have taken a backseat to shiny merchandise.”

Lower than per week later, one other staffer introduced their resignation on X, additionally citing safety concerns. One former staffer, Daniel Kokotajlo, mentioned in a Might 2024 weblog put up that he resigned as a result of he was “dropping confidence that it could behave responsibly across the time of AGI.”

Kokotajlo later advised Fortune that OpenAI initially had about 30 individuals researching questions of safety associated to AGI, a nonetheless theoretical model of AI that causes in addition to people, however a sequence of exits decreased that head depend by nearly half.

The corporate’s former head of preparedness, Aleksander Madry, assumed a brand new function in July 2024. The place is a part of OpenAI’s Security Methods group, which develops safeguards, frameworks, and evaluations for the corporate’s fashions. The job pays $555,000 a yr plus fairness.

“You may be the immediately accountable chief for constructing and coordinating functionality evaluations, menace fashions, and mitigations that kind a coherent, rigorous, and operationally scalable security pipeline,” the job itemizing says.