This video by Wes Roth, published in December 2025, discusses the release of OpenAI's GPT-5.2, describing it as a massive leap forward rather than a small incremental update. The second half of the video focuses on the economic implications, specifically analyzing a new benchmark called "GDP-eval," which measures performance on real-world, economically valuable tasks. In this benchmark, GPT-5.2 Pro achieved a 74% win/tie rate against human industry experts—a significant jump from the ~39% score of previous models just months prior. Roth argues this signals a critical turning point where AI is beginning to outperform experienced professionals (with an average of 14 years of experience) at a fraction of the cost, citing a 400x cost reduction in one year. The video concludes with a discussion on the potential for "catastrophic job loss" as AI intelligence per dollar continues to skyrocket, validating fears that human labor in many sectors could soon be replaced. (Gemini 3 Pro assisted with this summary).
