Anticipate and mitigate AI dangers

OpenAI has announced a call to hire a Preparedness Manager, a “crucial” position to continue the evolution of artificial intelligence (AI) models, preventing them from causing harm in the real world.

AI models are becoming more capable, but these capabilities pose real-world risks, such as mental health and security issues, with their capabilities abused to personalize malicious campaigns or even generate new malware.

The capabilities of these models also allow them to detect critical vulnerabilities in computer systems, which has led OpenAI, responsible for the GPT models and the ChatGPT ‘chatbot’, to look for a Preparedness manager.

This position is “crucial”, as the company’s CEO, Sam Altman, has stated in his X account, where he explained that “it will be a stressful job” and whoever ends up occupying it “will be involved in a complicated situation almost immediately.”

This is because their main duty will be to anticipate the dangers posed by AI in the real world to mitigate them, both biological and cybersecurity, as explained in the job offer.

To do this, you will have to “develop and coordinate capability assessments, threat models and mitigations that make up a coherent, rigorous and operationally scalable security process.”