OpenAI is recruiting high-salary security experts to address the risk of AI self-evolution.

OpenAI's goal is to develop an AI tool that can self-improve. Currently, the company is preparing for the risks that come with it. In the past six months, after significant progress with OpenAI and Anthropic's programming tools, the potential for AI systems to achieve "recursive self-improvement" has become a focus for industry leaders. Demis Hassabis stated this week that humanity is now standing at the "singularity" - the moment when AI begins to self-develop and surpass human intelligence. OpenAI, set to launch this year, recently posted an unusual job opening: the company is looking for a security researcher specialized in studying the consequences of "when AI can train better versions of itself". The job posting was released this month by OpenAI's Preparedness security team. The total annual compensation for the position ranges from 250,000 to 380,000 euros, targeting "experienced technical experts to support preparedness for recursive self-improvement". The job description states: "This role is built on thinking about problems that do not currently exist but may arise in the future." "Therefore, a cautious and strategic mindset is crucial for this position."