AI's Evolution Creates Hybrid Jobs and Sets New Benchmarks
Advanced AI systems are now outperforming previous generations on demanding 2023 benchmarks like MMMU and GPQA. This rapid evolution is also transforming the job market by creating “hybrid roles” that blend human oversight with machine efficiency, particularly in sectors like healthcare and finance. Experts agree that adaptability and digital literacy are becoming essential workforce skills.
The MMMU (Massive Multi-discipline Multimodal Understanding) benchmark evaluates AI on 11,500 college-level questions across 30 subjects, from Art & Design to Engineering. Even top models like GPT-4V initially found it challenging, achieving only 56% accuracy, highlighting the benchmark's difficulty. Newer versions like MMMU-Pro have been introduced to further test the limits of multimodal reasoning. GPQA (Graduate-Level Google-Proof Q&A) is a benchmark designed with questions so difficult that a person with unrestricted internet access can't easily find the answers. While PhD-level experts score around 65-74%, early models like GPT-4 were at 39%. However, performance has surged, with models like Gemini 3 Pro and Grok 4 now scoring 92% and 88% respectively, far surpassing human expert performance. In healthcare, AI is not replacing surgeons but augmenting their abilities. AI-assisted surgeons use robotic tools for enhanced precision in minimally invasive procedures and leverage AI to analyze medical images for pre-operative planning. This creates hybrid roles where medical expertise is combined with the ability to operate and interpret AI-driven surgical systems. New hybrid roles in finance include AI model auditors, compliance specialists, and AI-enhanced investment analysts. Financial institutions are increasingly using AI for tasks like fraud detection and data reconciliation, shifting human professionals to focus on interpreting AI-driven insights and making strategic decisions. JPMorgan Chase has even filed a trademark for "IndexGPT," an AI tool to help select financial securities. The rise of AI has led to a demand for new "hybrid skills." In finance, this means combining financial knowledge with an understanding of AI ethics, governance, and data analysis. The Swedish fintech company Klarna reported in 2024 that 87% of its employees use generative AI in their daily tasks across compliance, customer support, and legal operations. The World Economic Forum predicts that while AI could displace 83 million jobs by 2027, it will also create 69 million new ones. This results in a net displacement of 14 million jobs, primarily affecting roles with routine and repetitive tasks. However, a Goldman Sachs report estimated that generative AI could expose approximately 300 million full-time jobs worldwide to some degree of automation.
