Unreasonable Labs

Unreasonable Labs is developing an artificial intelligence platform aimed at accelerating scientific discovery across various fields. The company emerged from stealth in March 2026, securing a $13.5 million funding round led by Playground Global, with participation from AIX Ventures, E14 Fund, and MS&AD Ventures. This capital is intended to scale its core technology and expand its technical team.

The company was co-founded by CEO Yuan Cao and CTO Markus Buehler. Cao was formerly a Senior Staff Research Scientist at Google DeepMind, where he contributed to the Gemini model family. Buehler is an engineering professor at MIT, recognized for his work in AI-driven materials science and computational engineering. Their combined expertise unites elite AI research with deep physical engineering knowledge. The advisory board includes notable figures such as Nobel Prize-winning physicist Kostya Novoselov and Robert Langer, an MIT Institute Professor.

Unreasonable Labs' core technology addresses the limitations of current large language models (LLMs) in generating truly novel scientific discoveries. The platform integrates LLMs with a structured network of relationships and neurosymbolic mathematical abstractions. This approach transforms unstructured information from disparate scientific domains into a verifiable network, enabling the system to reason causally, synthesize new insights, and generate testable hypotheses. The goal is to move beyond simple data retrieval and summarization to compose new knowledge across disciplines like chemistry, materials science, and biology. The system is designed to support multiple stages of R&D by ideating across fields, simulating solutions, and helping to design physical experiments.

The business model involves providing this AI "operating system" to R&D teams in various industries. Unreasonable Labs has initiated pilot collaborations with industrial partners in the energy, materials science, and pharmaceutical sectors. The company operates from offices in Cambridge, Massachusetts, and Palo Alto, California.

Keywords: AI scientific discovery, neurosymbolic AI, knowledge discovery engine, computational science, R&D acceleration, materials science AI, generative discovery, cross-disciplinary research, hypothesis generation, AI for chemistry, AI for biology, research and development platform, human-led AI exploration, superintelligence, Playground Global, Yuan Cao, Markus Buehler