Andrew Tullock (Thinking Machines Lab): 10 Key Things You Must Know

Overview

Andrew Tulloch is a distinguished Australian computer scientist, machine learning expert, and AI entrepreneur notable for his contributions to pioneering artificial intelligence projects and his influential role in the AI startup ecosystem. Formerly a key figure at Meta (formerly Facebook) and OpenAI, Tulloch has made headlines recently for co-founding Thinking Machines Lab, a cutting-edge AI research and product company focused on customizable and interpretable AI systems. His refusal of a landmark $1.5 billion offer from Meta’s Mark Zuckerberg to return to the tech giant underscores his commitment to scientific independence and innovation beyond conventional corporate frameworks. This article explores ten compelling facts about Andrew Tulloch’s professional journey, technical achievements, and visionary impact on AI technology.

1. Early Academic and Professional Background

Andrew Tulloch hails from Australia and has a strong foundation in mathematics and statistics. He graduated with first-class honours and was awarded a University Medal in advanced mathematics from the University of Sydney. Subsequently, he obtained a Master’s degree with distinction in Mathematical Statistics from Trinity College, Cambridge University. This exceptional academic pedigree laid the groundwork for his career in complex machine learning problems and large-scale AI system design, equipping him with deep expertise in both mathematical theory and practical implementation.

2. Groundbreaking Work at Meta (2012–2023)

Tulloch spent over a decade at Meta, where he played an instrumental role in designing and building scalable machine learning systems. One of his hallmark contributions was to PyTorch, an open-source AI framework that has become one of the most widely adopted tools across the AI research and development community worldwide. His efforts fundamentally shaped Meta's AI infrastructure, influencing applications from advertising algorithms to core AI research platforms, earning him the reputation of a visionary engineer deeply invested in creating efficient, scalable AI tools.

3. Transition to OpenAI and GPT-4 Contributions

In 2023, Andrew Tulloch joined OpenAI, where he worked on critical pretraining and reasoning models for GPT-4 and GPT-4.5. His expertise in large-scale machine learning and distributed systems enabled significant advances in computer model training and natural language understanding. As a Member of Technical Staff, he contributed directly to cutting-edge AI capabilities that helped fuel the widespread adoption of generative AI models in multiple domains, solidifying his status as a key player in AI development.

4. Co-founding Thinking Machines Lab

Alongside Mira Murati, the former CTO of OpenAI, Tulloch co-founded Thinking Machines Lab in early 2025. The startup aims to develop AI systems that go beyond typical chatbot interfaces by focusing on safety, interpretability, and customization. Their vision is to democratize AI technology, making it more adaptable to individual users’ goals and broadly comprehensible. This initiative reflects Tulloch’s belief that AI must empower people, not just corporate interests, and serves as a platform for advancing AI research openness and innovation.

5. Resisting a $1.5 Billion Offer from Meta

In one of the most publicized stories of Silicon Valley’s AI talent battles, Andrew Tulloch declined a $1.5 billion, six-year compensation package personally offered by Meta CEO Mark Zuckerberg. This extraordinary offer was contingent on bonuses and stock performance and came in the wake of Meta’s unsuccessful attempt to acquire Thinking Machines Lab for over $1 billion. Tulloch’s decision to refuse the offer and remain committed to his startup demonstrates his dedication to nurturing independent AI research rather than returning to a corporate tech environment.

6. Driving Values of Scientific Freedom and Independence

Tulloch’s career decisions highlight a broader value system prioritizing scientific freedom, transparency, and collaboration in AI development. By turning down lucrative corporate recruitment and focusing on startups that emphasize community collaboration and open dissemination of research, he aligns with a growing movement in AI to democratize technology access and reduce centralized control over powerful AI tools.

7. Technical Expertise and Programming Skills

Throughout his career, Tulloch has demonstrated versatility with numerous programming languages and technical tools, including C++, Python, Rust, Haskell, Go, R, and MATLAB. His proficiency in these languages reflects his capability to develop complex machine learning algorithms, optimize large distributed systems, and engineer AI platforms that operate efficiently at scale.

8. Educational Pursuits and Doctoral Studies

In addition to his professional work, Andrew Tulloch is engaged in advanced academic research, currently undertaking doctoral studies in Statistics and Machine Learning at the University of California, Berkeley. This ongoing academic pursuit underscores his commitment to deepening theoretical understanding and contributing novel insights to the scientific foundations of AI.

9. Thinking Machines Lab’s Vision and Impact

Under Tulloch’s leadership, Thinking Machines Lab has secured substantial funding—reportedly over $2 billion—and is valued at over $12 billion despite not having launched a product yet as of mid-2025. The lab's mission focuses on building AI tools that are more transparent, customizable, and capable than existing systems, addressing important challenges of safety, explainability, and user control in AI technologies.

10. Influence on the Future of Artificial Intelligence

By spearheading innovative projects and rejecting lucrative corporate offers, Andrew Tulloch has become a symbol of the evolving AI landscape where research freedom and ethical concerns increasingly influence industry directions. His work at Thinking Machines Lab and previous contributions to PyTorch and GPT architectures are shaping next-generation AI platforms designed to be more adaptable and accountable, potentially transforming how AI integrates into society.

Conclusion

Andrew Tulloch stands out as a leading figure in the AI research community, renowned for his technical prowess, visionary leadership, and principled stance on scientific independence. From his foundational academic achievements to his milestone contributions at Meta and OpenAI, culminating in the co-founding of Thinking Machines Lab, Tulloch’s journey reflects the dynamic and high-stakes environment of contemporary AI development. His refusal of one of Silicon Valley’s most significant financial offers to maintain a path of innovation and autonomy highlights a shift in priorities among AI experts. As AI technologies continue to evolve rapidly, Tulloch’s efforts to create safer, customizable, and interpretable AI systems will remain imperative in balancing technical progress with societal values. Will his vision inspire a wider movement for openness and ethics in AI? Only time will tell.