A developer has created a live tracker to visualize the performance changes of major AI models over time, focusing on ELO ratings. They seek community input on historical datasets that reflect consumer experiences with AI models, as current benchmarks may not capture real-world usage accurately.