A surprising assertion about the future of jobs within artificial intelligence labs is igniting debate throughout the tech industry. Comments attributed to an OpenAI employee suggest that research positions may be automated by AI before engineering or sales roles. This claim implies that even highly intellectual jobs might not be as secure as previously thought as AI systems become increasingly capable. Researchers vs. AI The remarks were made by Hyperbolic Labs Co-Founder and CTO Yuchen Jin, who stated that an OpenAI researcher informed him that “researchers are considered the first kind of job that would be automated. ” Jin elaborated on X that infrastructure engineers would be next in line, while sales teams would remain human-led the longest. Although the OpenAI employee was unnamed, the quote rapidly went viral because it challenges the conventional belief about job security in AI research labs. Jin explained that much of modern research work involves tasks such as idea generation, designing experiments, testing various options, and analyzing results. These are areas where highly advanced AI systems can already operate swiftly and at scale. He added that most research processes follow predictable patterns, which makes them easier for AI to mimic.

However, he noted that top researchers—who consistently redefine problems—will still be difficult to replace. This may render infrastructure engineers more resilient to the complexity of AI systems. Jin characterized AI infrastructure as chaotic, full of edge cases, and built on highly customized codebases that AI models have not been trained on. Mistakes can be costly, requiring human judgment. Sales roles appear to be the safest for now, as they heavily depend on trust, relationships, emotions, and negotiation skills. Jin referred to sales as the “final boss” for AI.