IIT Madras Alumni Launch $1 Million Challenge for Materials Science

Two alumni from the Indian Institute of Technology (IIT) Madras have initiated a $1 million prize competition aimed at resolving five persistent problems in the field of materials science. This challenge is designed to leverage artificial intelligence to accelerate scientific discoveries and innovations.

Advaith Sridhar and Akash Ramdas, who became acquainted over a decade ago while studying at IIT Madras, co-founded Discovered Materials. The San Francisco-based startup focuses on AI-driven methods to identify new materials for semiconductor technology, as detailed on the company's website.

As reported by TechCrunch on November 5, the initiative has engaged academics from prestigious universities to discern five scientific issues that have proven resistant to solutions despite extensive research efforts. The subjects at hand involve complex areas such as cuprate superconductivity, the prediction of electrical properties in semiconductor interfaces, and the development of accelerated, more precise techniques for quantum chemistry calculations. The remaining challenges pertain to understanding how light influences materials and forecasting heat transfer across solid material interfaces.

Addressing these issues is vital, as breakthroughs could lead to enhanced electronic devices, advancements in semiconductor manufacturing, and improvements in energy technologies. Financially, the competition allocates a top prize of $400,000 for solutions pertaining to cuprate superconductivity. Other challenges feature rewards of $200,000 for two questions and $100,000 for the remaining two. The competition is accessible to individuals, teams, and organisations and does not have a prescribed deadline. Prizes will only be awarded if the proposed solutions fulfil the predetermined scientific criteria.

Prior to launching this competition, Discovered Materials experimented with sophisticated AI models on the identified problems. These models were permitted to operate as independent agents with access to online information and other research resources. However, the results highlighted the complexity of these questions, as the models made little headway after several days of autonomous work.

Sridhar posed the question, “How do we know AI is making progress?” underlining the challenges of quantifying scientific advancements in materials science. Stanford University professor Felipe Jornada, who assisted in formulating the challenges, remarked, “Across materials science, there are phenomena we observe but cannot explain.” Sridhar clarified that the competition is also aimed at developing benchmarks for evaluating the progress of both AI technologies and researchers when tackling difficult scientific challenges.

In August, the startup secured $9 million in seed funding to bolster its research into semiconductor materials. The road ahead could not only influence scientific progress but also reshape the landscape of materials research through innovative use of artificial intelligence.