The landscape of artificial intelligence is shifting rapidly. We are moving past the era where AI was primarily a passive tool for generating text or answering questions. Today, the industry is racing toward systems that can act, reason, and execute complex workflows autonomously. At the center of this evolution at OpenAI is Thibault Sottiaux, an engineer whose work has already reshaped how developers interact with code, and who is now spearheading one of the most significant overhauls in ChatGPT’s history.
From Codex to Code Interpreter: A Track Record of Growth
Sottiaux, often referred to as Tibo, has been a key architect behind some of OpenAI’s most successful ventures. His early work focused heavily on AI coding, a domain that has emerged as one of the company’s fastest-growing business segments. The development of Codex and subsequent coding tools demonstrated that large language models could do more than just chat; they could understand logic, write functional code, and assist developers in speeding up their workflows.
This success wasn’t just a technical milestone; it proved the commercial viability of AI in professional settings. By helping engineers build software faster and more efficiently, Sottiaux and his team helped establish a new standard for AI utility. The ability to translate natural language into executable code opened doors for “vibe coding” and low-barrier development, allowing non-technical users to create applications and automating tedious tasks for senior developers. This foundation in code generation provided the technical backbone for many of the advanced features we see today.
A New Mission: Overseeing the ChatGPT Overhaul
With a proven track record in building reliable, high-impact AI tools, Sottiaux has now taken on a broader mandate. He is currently overseeing a sweeping transformation of ChatGPT itself. This isn’t a minor update or a simple feature addition; it represents a fundamental rethinking of how the interface and the underlying model behavior interact with users.
The goal is to move ChatGPT beyond a conversational interface and turn it into a true agent capable of handling complex, multi-step tasks. This overhaul involves refining “model behavior” to ensure that the AI doesn’t just respond, but acts with precision, safety, and reliability. Sottiaux’s role is critical in bridging the gap between raw model capabilities and a user experience that feels seamless and trustworthy.
Redefining Model Behavior
One of the most challenging aspects of AI development is controlling how a model behaves in edge cases. As models become more powerful, ensuring they follow instructions accurately, avoid hallucinations, and maintain safety standards becomes paramount. Sottiaux’s work focuses on these nuances. The overhaul aims to make ChatGPT more deterministic in its actions, allowing it to use tools, browse the web, and execute code with a level of confidence that users can rely on for real-world tasks.
This shift is essential for the next generation of AI applications. Users don’t just want an AI that can write a poem; they want an AI that can research a topic, compile data, draft a report, and format it correctly without constant supervision. Sottiaux’s engineering leadership is driving the technical changes that make this level of autonomy possible.
What This Means for the Future of AI
The transformation led by Sottiaux signals a broader trend in the AI industry. We are seeing a move toward “agentic” workflows, where AI systems take ownership of entire projects rather than waiting for step-by-step prompts. This has profound implications for productivity, creativity, and how we interact with technology in our daily lives.
For OpenAI, this overhaul is a strategic necessity. As competition heats up with rivals launching their own advanced agents and coding assistants, ChatGPT needs to evolve to maintain its lead. By leveraging the lessons learned from the success of AI coding, Sottiaux is helping to build a ChatGPT that is not only smarter but also more capable of taking action in the real world.
Thibault Sottiaux’s journey from leading the charge on AI coding to steering the massive transformation of ChatGPT highlights the dynamic nature of work at OpenAI. It also underscores a key truth about the future of AI: the most valuable models will be those that can behave reliably, execute complex tasks, and integrate seamlessly into our workflows. With Sottiaux at the helm, the next chapter of ChatGPT promises to be one of the most significant yet.
