Duke University
Boris PowerHead of Applied Research · OpenAIResearcher↗
Researcher · Founder · Executive · Engineer
Noam Shazeer
Shazeer helped create the Transformer, the design behind modern language models. He also pioneered systems that activate only the parts of a model needed for each task, making very large models practical. He founded Character.AI and joined OpenAI in 2026 after returning to Google to co-lead Gemini.[1]
Career & education
Current role
OpenAIJune 2026-present
OpenAIJune 2026-presentPreviously
Google DeepMind2024-June 2026VP Engineering, Gemini co-lead
Character.AI2021-2024Co-founder and CEO
Google2012-2021Principal Software Engineer
Google DeepMind2024-June 2026Character.AI2021-2024Google2012-2021Education
Work and education history is primarily focused on AI-relevant roles and may not be comprehensive.
Citation Trend
Total citations to scholarly work, including research papers, technical reports, books and theses, as recorded by Google Scholar. Each point is a dated snapshot.
citations411,524As of Nov 1, '26
citations+11,011Since last snapshot, Oct 1, '26
Updates 1
Selected contributions
Attention Is All You Need
With Ashish Vaswani, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan Gomez, Lukasz Kaiser and Illia PolosukhinThe Transformer uses attention to translate text while allowing faster training.
Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer
With Colin Raffel, Adam Roberts, Katherine Lee, Sharan Narang, Michael Matena, Yanqi Zhou, Wei Li and others ·T5 treats language tasks as text input and text output within one framework.
PaLM: Scaling Language Modeling with Pathways
With Aakanksha Chowdhery, Sharan Narang, Jacob Devlin, Maarten Bosma, Gaurav Mishra, Adam Roberts, Paul Barham and others ·Presented PaLM, a large language model demonstrating gains in reasoning, translation and coding.
References
- Axios: Shazeer’s move to OpenAIaxios.com
June 2026 move from Google’s Gemini leadership to OpenAI after founding Character.AI.
- Fast Transformer Decoding: One Write-Head is All You Needarxiv.org
He also introduced multi-query attention, which reduces the memory needed to generate text and speeds up responses.
- GLU Variants Improve Transformerarxiv.org
His work on SwiGLU changed how information flows inside Transformers, improving model quality without increasing the amount of computation in the paper’s comparisons.
Based on public sources. If accuracy is important, verify details using the links or references.
For corrections, updates, or requests to remove information, contact us.
Reviewed September 5, 2026 · Checked for accuracy and new information.
