Mira Murati's Thinking Machines Lab ships its first product, and it is not a chatbot

Thinking Machines Lab, the research company founded by former OpenAI chief technology officer Mira Murati after her high-profile 2024 departure, has shipped its first commercial offering - not a consumer-facing chatbot competing head-on with ChatGPT or Claude, but a developer-focused platform for customising and fine-tuning open-weight models with what the company describes as substantially better sample efficiency and more transparent training diagnostics than existing tools. The choice to enter the market from the developer-tooling angle rather than the crowded consumer-assistant category reflects a deliberate positioning decision by a founding team stocked with researchers who spent years building the infrastructure underneath OpenAI's own model-training pipeline rather than the consumer product itself. The company's roughly two-billion-dollar seed and follow-on funding, raised well before any product existed, was justified to investors largely on the strength of Murati's leadership credibility and the research pedigree of a founding team that includes several other senior OpenAI alumni, mirroring the pure-reputation-based funding dynamic that has also characterised Safe Superintelligence's fundraising, and raising similar questions about whether such valuations are sustainable if the eventual product does not achieve rapid commercial traction. The fine-tuning and customisation tooling market that Thinking Machines has entered is already contested, with Together AI, Fireworks AI and the open-source Hugging Face ecosystem all offering comparable capability at varying degrees of maturity, meaning Thinking Machines will need to demonstrate a clear technical edge rather than relying on founder reputation alone to win developer mindshare and, eventually, enterprise contracts. For India's growing base of AI application developers building on open-weight models - drawn to open weights by cost sensitivity and data-residency preferences - better fine-tuning tooling addresses a real pain point, and several Indian AI-infrastructure startups have already begun evaluating Thinking Machines' platform as a potential complement to or replacement for existing fine-tuning workflows built on open-source tooling. What to watch: how the initial product's technical claims hold up under independent developer benchmarking, whether Thinking Machines follows its developer-tooling entry with a more direct foundation-model release, and whether the company's valuation trajectory tracks or diverges from Safe Superintelligence's as both attempt to convert founder reputation into durable product businesses.

Original source: TechCrunch