PeopleofSuperintelligence

Builder

James Molloy

James Molloy co-led infrastructure for Gemini 1.0, a multimodal model family reported in 2023. His earlier work includes the 2019 Nature paper Grandmaster level in StarCraft II, which uses multi-agent reinforcement learning, and the 2021 arXiv preprint Automap. In 2024 he co-authored PartIR, a partitioning system for neural networks, and the Gemini 1.5 report.

Timeline

  1. 2019

    Grandmaster level in StarCraft II

    He co-authors the Nature paper Grandmaster level in StarCraft II using multi-agent reinforcement learning, which appears in October 2019.

  2. 2021

    Automap for automated parallelism

    In 2021 he co-authors the arXiv preprint Automap: Towards Ergonomic Automated Parallelism for ML Models.

  3. 2023

    Infrastructure co-lead on Gemini 1.0

    He is infrastructure co-lead for Gemini 1.0, a multimodal model family whose report comes out in 2023.

  4. 2024

    The PartIR partitioning system

    He co-authors PartIR, a partitioning system for neural networks, whose arXiv version 4 is dated November 2024.

  5. 2024

    Gemini 1.5 report co-author

    He co-authors the Gemini 1.5 report, whose arXiv version is dated December 2024.

Sources and credits

Photo: We have no published photo of James Molloy yet. Send us a free-licensed photo.

  1. Grandmaster level in StarCraft II using multi-agent reinforcement learning | Nature nature.com. Checked 14 October 2026.
  2. [2112.02958] Automap: Towards Ergonomic Automated Parallelism for ML Models arxiv.org. Checked 14 October 2026.
  3. PartIR: Composing SPMD Partitioning Strategies for Machine Learning arxiv.org. Checked 14 October 2026.
  4. Gemini 1.5: Unlocking multimodal understanding across millions of tokens of context arxiv.org. Checked 14 October 2026.
  5. Gemini: A Family of Highly Capable Multimodal Models arxiv.org. Checked 14 October 2026.
  6. Gemini - A Family of Highly Capable Multimodal Models: Contributions and Acknowledgments | HackerNoon hackernoon.com. Checked 14 October 2026.
  7. [2401.11202] PartIR: Composing SPMD Partitioning Strategies for Machine Learning arxiv.org. Checked 14 October 2026.
  8. [2403.05530] Gemini 1.5: Unlocking multimodal understanding across millions of tokens of context arxiv.org. Checked 14 October 2026.