Thinking Machines Lab: The next big AI Research and Product Company?
Created on February 19|Last edited on February 19
Comment
Thinking Machines Lab is a newly launched artificial intelligence research and product company focused on making AI more widely understood, customizable, and capable. Despite rapid advancements in AI capabilities, many challenges remain. The scientific community struggles to keep pace with cutting-edge developments, training methodologies are concentrated within a few top research labs, and current AI systems are difficult to personalize. Thinking Machines Lab aims to address these gaps by building AI that is more flexible, adaptable, and accessible.
A team with deep AI expertise
The current founding team at Thinking Machines Lab includes 29 researchers, engineers, and product builders, with 21 of them previously working at OpenAI (according to my count). Their collective experience includes developing some of the most widely used AI models and platforms, such as ChatGPT, Character.ai, and Mistral, as well as open-source tools like PyTorch, OpenAI Gym, Fairseq, and Segment Anything. Mira Murati, former CTO of OpenAI, now leads the company as CEO, while John Schulman, co-creator of reinforcement learning techniques that power modern AI chatbots, serves as Chief Scientist.
Commitment to "open" science and collaboration
Thinking Machines Lab is committed to open scientific progress. The company plans to publish technical blog posts, research papers, and open-source code to advance the broader AI research community. By openly sharing insights and methods, they hope to contribute to a deeper understanding of AI while also improving their own research culture.
Human-AI collaboration over full automation
Rather than focusing purely on autonomous AI, Thinking Machines Lab emphasizes human-AI collaboration. Their approach involves developing multimodal systems that enhance human capabilities rather than replace them. They aim to create AI that adapts to users' expertise across various fields, from programming and science to creative and professional applications.
Building strong AI foundations
Beyond adaptability and collaboration, Thinking Machines Lab prioritizes building intelligent models that push the boundaries of AI capabilities. They focus on foundational model intelligence, recognizing that the most advanced AI models can drive breakthroughs in science, engineering, and other fields. Infrastructure quality is also a key concern, as they aim to develop highly efficient and reliable AI systems that enhance research productivity without cutting corners.
Advancing AI Safety Through Practical Testing
AI safety is another core pillar of the company’s mission. Thinking Machines Lab plans to take an empirical and iterative approach to safety, combining proactive research with real-world testing. Their safety strategy includes preventing misuse while maintaining user freedom, sharing best practices for building secure AI, and supporting external alignment research through open code and datasets. The company sees present-day safety methods, such as red-teaming and post-deployment monitoring, as essential for guiding the safe development of future AI systems.
A New AI Powerhouse in the Making?
With a leadership team largely drawn from OpenAI and a clear focus on AI research, safety, and product development, Thinking Machines Lab is positioning itself as a major new player in the AI space. By prioritizing collaboration, adaptability, and scientific openness, the company is taking a deliberate approach to building AI that serves a wide range of users and applications.
Add a comment
Tags: ML News
Iterate on AI agents and models faster. Try Weights & Biases today.