Scaling laws: Large Language Models Are Improving Exponentially, doubling every 7 months

Amazing stuff!

Dozens of research articles have been written on the scaling laws of machine learning!

"... According to a metric it devised, the capabilities of key LLMs are doubling every seven months. This realization leads to a second conclusion, equally stunning: 

By 2030, the most advanced LLMs should be able to complete, with 50 percent reliability, a software-based task that takes humans a full month of 40-hour workweeks. And the LLMs would likely be able to do many of these tasks much more quickly than humans, taking only days, or even just hours. ..."

Large Language Model Performance Raises Stakes - IEEE Spectrum In a few years, AI could handle complex tasks with ease

METR: Model Evaluation & Threat Research

Measuring AI Ability to Complete Long Tasks (open access)