Announced 5 Nov 2026 · Sources checked
A $1 million challenge for AI materials science
Discovered Materials, a months-old San Francisco startup using AI models to find new substances for building advanced chips, announced a million-dollar challenge for AI models to solve five critical problems in materials science.
The company is offering a pool of prizes worth $1 million in total for solutions to the problems. The challenges were written with an advisory group of academics at leading global universities, recruited after the founders asked what the grand challenges in the field were and how anyone would know AI is making progress.
It is more difficult to specify problems clearly in materials science because it is not like math, where you have a very specific definition and can get to a yes or no, Advaith Sridhar, a co-founder of Discovered Materials, told TechCrunch. We came to this discussion wondering what the grand challenges are in this field and how we know AI is making progress.
The five problems
The five open problems span theory and computation. The list includes a theory to explain cuprate superconductivity, the ability of certain materials to have zero electrical resistance at ultra-low temperatures.
Two of the problems are numerical methods aimed at the semiconductor industry directly: predicting the structures of interfaces between silicon and silicon dioxide, a material combination found in almost all semiconductors, and mapping grain structure to thermal properties to enable better heat dissipation materials.
The remaining problems are a theory to predict why shining light on materials can change their behavior, particularly around superconductivity, and a numerical method to improve Density Functional Theory, a key technique used in materials science to model the structures of atoms and molecules.
Across materials science there are phenomena we observe but cannot explain, Felipe Jornada, a Stanford professor who helped develop the challenges, said in a statement. Explaining these phenomena would improve our understanding of physics, and the answers would directly shape how we build electronics, store energy, and manufacture the materials of the future.
Frontier models already tried, and mostly failed
Before announcing the challenges, Discovered Materials turned frontier lab models loose on them, giving AI agents internet access and other resources to work autonomously for days. The models spent several days on each problem without making much headway, according to the company.
The results are a useful baseline for the field. While frontier AI models have solved a remarkable series of outstanding mathematics problems and made progress in the life sciences, they have yet to truly plumb the complex physics in advanced materials.
The company says it does not care who solves the problems first. It is not particularly a concern to us whether we find it, or someone else finds it, or a professor finds it, or the frontier labs find it, Sridhar said. The reason we are putting out these challenges is because we want them to serve as the goalposts.
If someone goes and solves these challenges, Sridhar added, the field is deepening its understanding of materials and how things around us behave. Tracking efforts to solve the problems will show whether the models are climbing in this domain.
Why a startup would set science's goalposts
The challenge reflects the changing nature of scientific inquiry as AI proves capable of uncovering insights and techniques human researchers did not anticipate. With solutions apparently easier to find than ever, the questions become more important.
This is a new way of thinking about science, where we come up with the verification and the challenges first, and then we let somebody come up with the ideas that explain the problem, Sridhar said. Knowing what are the right questions to ask is not trivial and very difficult in this domain, so this is where the professors have really helped us shape the right questions.
Materials science does not have an equivalent of the Millennium Prize Problems, the famous set of mathematics problems with million-dollar rewards. Discovered Materials is effectively trying to build that equivalent for its own field, with AI systems as the intended solvers.
The company behind the prize
Discovered Materials is part of Y Combinator's spring 2026 batch and launched in August with a $9 million seed round led by Lightspeed India Partners, according to prior coverage. Its main focus is developing new materials for use in semiconductors, and it partners with chip giants on research projects while also funding its own scientific efforts.
When the company announced its seed round in August, Sridhar told TechCrunch he hoped the company would generate a useful material within a year. Those timelines have accelerated a little bit, he says today. He said he is more confident now that the company will find industrially useful materials on a timeline shorter than a year.
The startup's research program combines AI agents with materials science software tooling. It says it makes best-effort attempts to experimentally validate discovered materials in its own lab, and it publishes benchmark work on a material discovery bench linking standard materials science packages.
The $1 million challenge is measured against the company's own scale: a seed-stage startup is funding prizes that rival the Millennium Problems' individual payouts, an aggressive bet that the bottleneck in AI-driven materials discovery is the questions, not the models.
What comes next
It is easy to imagine OpenAI or Anthropic attempting to one-shot the Discovered Materials problems with significantly more compute or a newer model, as TechCrunch noted. The five problems are public, so any lab can attempt them.
For materials science, the challenge establishes a public benchmark where there was none. Whether the $1 million goes to a frontier lab, an academic group, or stands unclaimed, the problems themselves now serve as a measurable yardstick for whether AI is actually making progress in the physical sciences.
For Discovered Materials, the prize is also a recruiting and credibility signal. A company three months out of Y Combinator is defining the verification criteria for its own field, and if any solver succeeds, the startup gets credit for asking the right questions even if it does not produce the material itself.
Common questions
Who is behind the $1 million materials science challenge?
Discovered Materials, a San Francisco startup from Y Combinator's spring 2026 batch, is offering the prize pool. The company uses AI models to discover new materials for semiconductors and announced a $9 million seed round led by Lightspeed India Partners in August 2026.
What are the five unsolved problems in the challenge?
They are: a theory of cuprate superconductivity, a numerical method for silicon-silicon dioxide interfaces, a numerical method mapping grain structure to thermal properties, a theory explaining light-induced changes in materials, and a numerical method to improve Density Functional Theory.
How did the frontier AI models do on the problems?
Before announcing the challenge, Discovered Materials gave frontier lab models internet access and other resources to work autonomously for days on each problem. The models spent several days per problem without making much headway.
Does the winner need to be an AI system?
The intended solvers are AI models, but the company said it does not care who solves the problems first, whether it is the company itself, another lab, or a professor. The point is to establish measurable goalposts for AI progress in materials science.
Sources & further reading
How this story was made
Written by Kristian Kostov with AI assistance and checked against the linked sources. Company performance claims are attributed to the company. Analysis reflects AiLookout’s interpretation; we have not independently tested the products discussed. The cover is an AI-generated editorial illustration, not a photograph or evidence of an actual event.
Our editorial standards
