PeopleofSuperintelligence

Lab leader

Anna Goldie

Also known as Anna Darling Goldie

Anna Goldie is a computer scientist and AI lab founder whose work connects machine learning with the hardware that runs AI. At MIT, she built a Mandarin dialogue system and studied conversational language generation. Her reinforcement-learning research moved from placing neural-network operations across devices to AlphaChip's chip floorplanning. She co-authored Constitutional AI and completed a Stanford PhD on reinforcement learning, chip design, and language modeling. TIME named her to its 2026 TIME100 AI list; as of October 2026, she is founder and CEO of Ricursive Intelligence.

Now at
Ricursive Intelligence
Works
40

Publication numbers from OpenAlex, as of 14 October 2026. Citations and h-index are hidden because this OpenAlex record may be incomplete or mix in other authors.

Timeline

  1. 2011–2017

    A chatbot in Mandarin

    At MIT, she builds CHATTER, a spoken dialogue system for Mandarin-language learning, for her 2011 MEng thesis. A 2017 paper with five co-authors studies sequence-to-sequence models that generate longer, more informative responses to conversations.

  2. 2018

    Placing neural networks

    A 2018 paper introduces a hierarchical model that uses policy gradients to assign operations in computational graphs to hardware devices. It reports optimized placements for TensorFlow graphs with more than 80,000 operations.

  3. 2021

    Learning to lay out chips

    With Azalia Mirhoseini, Goldie co-led AlphaChip, a deep reinforcement-learning project for chip floorplanning. The Nature paper reports that its method was used to design Google's next-generation AI accelerators.

  4. December 2022

    Training harmless assistants

    Goldie co-authors Constitutional AI: Harmlessness from AI Feedback, which tests training a harmless assistant with written principles and AI preferences.

  5. June 2025

    A Stanford doctorate

    She completes a Stanford computer-science PhD focused on reinforcement learning for sequential decisions, with a dissertation spanning chip design and language modeling.

  6. December 2025

    A lab for chip design

    Goldie and Azalia Mirhoseini launch Ricursive Intelligence, a frontier AI lab focused on semiconductor design. Earlier, she co-founded Google Brain's ML for Systems team and worked at Anthropic and Google DeepMind.

  7. 2026

    A place on TIME100 AI

    TIME includes Goldie and Mirhoseini in its 2026 TIME100 AI list as co-founders of Ricursive Intelligence.

  8. As of October 2026

    Leading Ricursive

    She is the founder and CEO of Ricursive Intelligence, a frontier AI lab for chip design.

Sources and credits

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

  1. Anna Goldie annagoldie.com. Checked 14 October 2026.
  2. Generating High-Quality and Informative Conversation Responses with Sequence-to-Sequence Models arxiv.org. Checked 14 October 2026.
  3. Hierarchical Planning for Device Placement research.google. Checked 14 October 2026.
  4. A graph placement methodology for fast chip design nature.com. Checked 14 October 2026.
  5. Constitutional AI: Harmlessness from AI Feedback arxiv.org. Checked 14 October 2026.
  6. Christopher Manning and Ph.D. Students' Dissertations www-nlp.stanford.edu. Checked 14 October 2026.
  7. Ricursive Intelligence Launches Frontier AI Lab to Transform Semiconductor Design and Accelerate Path Toward Artificial Superintelligence prnewswire.com. Checked 14 October 2026.
  8. The 100 Most Influential People in AI 2026 time.com. Checked 14 October 2026.
  9. TechCrunch Disrupt 2026: Ricursive Intelligence’s Anna Goldie and Azalia Mirhoseini on when AI starts designing its own hardware techcrunch.com. Checked 14 October 2026.