Mira Murati’s Thinking Machines Acquires Workshop Labs — The AI Startup That Wants to Make Humans Irreplaceable
Summary
Workshop Labs, the research startup born from an essay series about AI taking everyone’s jobs, has been acquired by Mira Murati’s Thinking Machines. The deal unites two teams that share an uncommon conviction in the AI world: that the technology should make people matter more, not less.
Workshop Labs announced on April 13, 2026, that it is joining Thinking Machines, the AI lab founded by former OpenAI CTO Mira Murati. The acquisition brings together a small but technically accomplished team with one of the most closely watched new entrants in the AI industry.
Workshop Labs was founded by Rudolf Laine and Luke Drago, growing out of “The Intelligence Curse,” an essay series examining the economic consequences of transformative AI. The central argument was stark: if AI automates people out of the economy, it leaves them disempowered. Rather than accept that outcome, the founders built a startup focused on personalizing AI systems to each user’s unique knowledge, taste, and values — aiming to decentralize control over AI rather than concentrate it.
Despite being small and resource-constrained, the team built what it described as the first integrated private post-training and inference stack, achieved best-in-class training speeds for trillion-parameter models, and shipped a product that allowed anyone to tailor a model with their own data in a few clicks.
The connection with Thinking Machines developed organically. Murati reviewed drafts of the acquisition announcement, and Workshop Labs noted that “The Intelligence Curse” had been circulated internally at Thinking Machines to help new team members understand the lab’s mission. Thinking Machines co-founder John had been giving talks on keeping humans competitive as early as 2023, and Murati has consistently emphasized building AI that works alongside people rather than replacing them.
For Thinking Machines, the deal adds specialized expertise in personalization, user-aligned fine-tuning, and private inference — capabilities that directly support a product vision centered on individual empowerment rather than pure automation. Workshop Labs received early access to Tinker, Thinking Machines’ post-training system, and the two teams had been sharing research ideas on parallel tracks before deciding to merge.
The acquisition reflects a growing philosophical divide within the AI industry. While most major labs optimize for raw capability and enterprise scale, a smaller contingent is building around the premise that the most valuable AI systems will be those shaped by and accountable to individual users. Whether that thesis can compete commercially against the platform-scale approaches of OpenAI, Google, and Anthropic remains an open question — but Thinking Machines now has a deeper bench to test it.