# Ioannis Antonoglou

> Greek AI researcher and engineer Ioannis Antonoglou is co-founder, president and CTO of Reflection AI. A founding-era DeepMind engineer, he contributed to DQN, AlphaGo, AlphaGo Zero, AlphaZero and MuZero, then led reinforcement learning from human feedback work for Gemini. Reflection AI is pursuing open-weight models and autonomous agents.

- **Role:** Co-founder, President and CTO at Reflection AI
- **Organizations:** Reflection AI, Google DeepMind
- **From:** Thessaloniki, Greece
- **Nationality:** Greek
- **Education:** Electrical engineering, Aristotle University of Thessaloniki, Postgraduate studies in artificial intelligence, University of Edinburgh, PhD, Learning to Search in Reinforcement Learning, University College London
- **Known for:** Contributed to DeepMind research behind DQN, AlphaGo, AlphaGo Zero, AlphaZero and MuZero., Was part of the AlphaGo team when it defeated Lee Sedol 4-1 in 2016., Led reinforcement learning from human feedback work for Gemini at DeepMind.

## Career timeline

- **1988** — Born in Thessaloniki, Greece.
- **2012** — Joined the then-small DeepMind team in London.
- **2015** — Co-authored research on deep Q-networks.
- **2016** — Contributed to AlphaGo, which defeated Lee Sedol 4-1.
- **2017** — Co-authored AlphaGo Zero and AlphaZero research.
- **2019-2020** — Co-authored MuZero research, later published in Nature.
- **2023** — Completed a UCL PhD thesis on learning to search in reinforcement learning.
- **2024** — Co-founded Reflection AI with Misha Laskin in March.
- **2025** — Reflection announced a $2 billion raise and a shift toward frontier open intelligence.
- **2026** — Discussed open-weight agents and sovereign AI in public interviews.

## Achievements

- Contributed to DeepMind research behind DQN, AlphaGo, AlphaGo Zero, AlphaZero and MuZero.
- Was part of the AlphaGo team when it defeated Lee Sedol 4-1 in 2016.
- Led reinforcement learning from human feedback work for Gemini at DeepMind.
- Completed a PhD at University College London in 2023.
- Co-founded Reflection AI in 2024.

## Latest updates

- **2026-06** — At Panathenea 2026, discussed sovereign AI and the ability for countries and organizations to control and adapt their own models.
- **2026-02** — Explained Reflection AI’s open-weight strategy and its move to train frontier base models from scratch in a Turing Post interview.
- **2025-10** — Reflection AI announced a $2 billion raise and its frontier open intelligence strategy.

## Links

- LinkedIn: https://www.linkedin.com/in/ioannis-alexandros-antonoglou-45393253
- Twitter/X: https://x.com/real_ioannis

---

Profile page: https://yespress.io/ioannis-antonoglou
Published by YesPress — https://yespress.io
Last updated: 2026-09-29