Jaime Sevilla

Epoch AI

prof_pic.jpg

I am working towards a better understanding and management of Artificial Intelligence. Background in Mathematics and Computer Science.

I study trends in Machine Learning at Epoch AI.
I aim to deliver a scientific understanding of what is driving innovation in AI, and sketch the path of the technology in future decades.

I also strive towards better management of global risk through my role as advisor of the Observatorio de Riesgos Catastróficos Globales.

Learn more about my work in ‘The Researcher Trying to Glimpse the Future of AI’, by Will Henshall of TIME.

selected publications

  1. scale2030.png
    Can AI Scaling Continue Through 2030?
    Jaime Sevilla, Tamay Besiroglu, Ben Cottier, and 4 more authors
    2024
    Accessed: 2024-10-10
  2. GATE: An Integrated Assessment Model for AI Automation
    Ege Erdil, Andrei Potlogea, Tamay Besiroglu, and 6 more authors
    2024
  3. FrontierMath: A Benchmark for Evaluating Advanced Mathematical Reasoning in AI
    Elliot Glazer, Ege Erdil, Tamay Besiroglu, and 21 more authors
    2024
  4. data2024.png
    Will we run out of data? Limits of LLM scaling based on human-generated data
    Pablo Villalobos, Anson Ho, Jaime Sevilla, and 3 more authors
    2024
  5. compute-growth2024.png
    Training Compute of Frontier AI Models Grows by 4-5x per Year
    Jaime Sevilla, and Edu Roldán
    2024
  6. algorithmic-progress.png
    Algorithmic progress in language models
    Anson Ho, Tamay Besiroglu, Ege Erdil, and 6 more authors
    2024
  7. pcd.png
    Parameter, Compute and Data Trends in Machine Learning
    Epoch
    2022
  8. please.png
    Please Report Your Compute
    J. Sevilla, A. Ho, and T. Besiroglu
    Communications of the ACM, 2023
  9. compute-trends.png