Physics-Based AI Revolutionizes Drug Discovery, Says SandboxAQ COO

At the World Economic Forum in Davos, Andrew McLaughlin, the chief operating officer of SandboxAQ, announced a pivotal shift in artificial intelligence (AI) driven by physics-based simulations. He emphasized that this approach is set to significantly enhance various fields, including drug discovery, materials science, and chemistry.

According to McLaughlin, traditional large language models (LLMs) face inherent limitations, particularly in generating accurate information. Users have reported issues such as “hallucinations,” where models produce incorrect data with unwarranted confidence. He stressed the need to transition towards systems grounded in physical equations.

“There’s a whole new set of architectures, which we call quantitative AI,” McLaughlin explained during an interview. “These models leverage the power of AI to generate quick and precise answers while ensuring outputs are scientifically accurate and rigorous.” He noted that such models are better suited for applications that conventional language models cannot address effectively.

McLaughlin, who previously served as the deputy chief technology officer under President Barack Obama, highlighted that the most impactful applications of AI will emerge in biology, human health, and materials science. He pointed out a growing interest from pharmaceutical companies in adopting physics-based AI, stating, “Instead of just doing little proof of concepts, we’re actually signing major deals that reflect the fact that pharmaceutical and biotech companies are achieving meaningful results now.”

The California-based SandboxAQ provides AI and quantum solutions to a range of industries, including banking, biopharma, and government. The company, which is supported by Nvidia, secured over $300 million in its last funding round in December 2024, achieving a valuation of $5.6 billion.

Potential of Physics-Based AI in India

Discussing the opportunities in India, McLaughlin expressed enthusiasm about the country’s potential in physics-based AI, citing its robust talent pool in mathematics, physics, chemistry, and biology. “The scientific depth in India is unprecedented, unparalleled,” he remarked. However, he cautioned that this talent must be coupled with advanced computational skills, data center infrastructure, and industry partnerships to fully realize the potential.

McLaughlin underscored that India has the opportunity to emerge as a global leader in AI and quantum technology, particularly after missing earlier waves of technological advancement.

He also addressed the current investment landscape in AI, noting that while some players are reaping substantial returns, there is a pressing need for broader funding across a diverse array of AI strategies. “Deep tech requires a long time horizon and a significant capital investment long before it yields returns,” he stated.

McLaughlin advocates for public-private partnerships to stimulate larger investments in deep tech, suggesting that government involvement can catalyze growth in this critical area. This collaborative approach could help in nurturing the next generation of technologies that rely on physics-based innovations.

As AI continues to evolve, the integration of physics into its framework may well signify the next frontier in the quest for more accurate and effective solutions across various sectors.