Discovered Materials, a months-old San Francisco startup that uses AI models to find new substances for advanced chips, has announced a “million-dollar challenge” for AI models to solve five open materials-science problems, according to TechCrunch, which reported on November 5, 2026. The company was part of Y Combinator’s spring 2026 batch, and the prize pool totals $1 million.

The five problems

  1. A theory explaining cuprate superconductivity, the behavior of materials that conduct with no electrical resistance at ultra-low temperatures.
  2. A numerical method for predicting the structure of interfaces between silicon and silicon dioxide.
  3. A numerical method linking a material’s grain structure to its thermal properties, to help design better heat-dissipating materials.
  4. A theory predicting why shining light on a material can change its behavior, particularly around superconductivity.
  5. A numerical method for improving density functional theory, a technique for modeling atomic and molecular structures.

What is not stated

The article does not state eligibility rules, submission requirements, judging criteria or deadlines. It links to the company’s challenge page, which the reporter could not access.

Background

  • TechCrunch names Advaith Sridhar as a co-founder; a photo caption names Sridhar and Akash Ramdas as founders. The company announced a seed round in August 2026, but the article gives no amount.
  • It partners with chip companies on research projects and also funds its own scientific work.
  • The company recruited academics at leading global universities to define the problems, and Stanford professor Felipe Jornada helped develop them.
  • Before announcing the challenges, it gave frontier AI models internet access and other resources to work autonomously for days on each problem. According to the article, the models made little progress.

What was said

  • Sridhar on defining the problems: “It’s more difficult to specify problems clearly in materials science because it’s not like math.” On the challenges’ purpose: “we want them to serve as the goalposts.”
  • Jornada, in a statement: “Across materials science, there are phenomena we observe but cannot explain.”
  • Sridhar said in August he hoped the company would produce a useful material within a year. He now says he is more confident it will find industrially useful materials in less than a year: “Those timelines have accelerated a little bit.”

Sources and notes

This post is based on TechCrunch’s report by Tim Fernholz (5:00 a.m. PST, November 5, 2026), which draws on interviews with Sridhar and a statement from Jornada. Lens did not read the company’s challenge page or contact it.

Image: laboratory test tubes in a rack, with a flask and beakers filled with coloured solutions, by Belikov Maxim via Wikimedia Commons, licensed under CC BY 4.0; cropped. Stand-in photo; it is not related to Discovered Materials and does not show the events described.