The next step for scientific discovery: Merging AI and quantum

Artificial intelligence (AI) has exhibited an immense impact on scientific discovery in recent years. Materials science and structural biology are two fields that have particularly experienced the benefits of the technology, with tech giants and startups alike leveraging AI to streamline the production of novel materials, identification of protein structures, and length of discovery cycles.

Yet, beneath the excitement, a critical truth is often overlooked; the future of discovery will not be driven by AI alone. As it stands, AI possesses a number of scientific limitations, including an inability to distinguish between novel discovery and rediscovery, data leakage, and limited explainability.

The question that many researchers have found themselves asking is: how do we overcome these hinderances while continuing to leverage the power of AI? The answer lies in the integration of AI into existing scientific processes with realistic expectations and rigorous validation procedures. First principles modeling, a scientific approach grounded in quantum mechanics, has emerged as a primary candidate for achieving this outcome while also paving the way for scientists to access the potential of AI and quantum computing.

Quantum and AI are enhancing our knowledge-based understanding of the world by providing powerful tools to analyze complex data, uncover hidden patterns, and drive scientific discoveries. – Julian van Velzen

Why quantum matters

Scientific innovation revolves around the laws of physics. While AI is an optimal tool for recognizing patterns amidst vast datasets, it lacks the ability to understand scientific behaviors in real-world environments.

First principles modeling, also referred to as ab initio modeling, uses quantum mechanics to more accurately predict material properties and molecular behavior. Not only does this framework ensure a more reliable foundation for experimentation by relying on physics, it also introduces the following four key benefits:

1. Validation of AI predictions: Quantum mechanical methods like the Density Functional Theory (DFT) and Coupled-Cluster theory can confirm provide a physics-based validation of AI-generated candidates before having to invest in costly experiments.
2. Exploration of novel domains: Quantum mechanics can be used to model systems where limited or no training data already exists, opening pathways to true innovation.
3. High-quality training data: Simulations informed by quantum mechanics help establish robust datasets that improve AI accuracy over time.
4. Mechanistic understanding: Unlike black-box AI predictions, quantum methods explain why a material or molecule behaves as it does, enabling smarter, more informed experimentation.

In using quantum mechanics as the foundation for discovery, first principles modeling provides an increased level of reliability that AI can’t deliver on its own. Quantum computing promises to accelerate these benefits even further by making experimental processes faster and more efficient, significantly expanding upon the boundaries of what’s scientifically possible.

The combination of AI and quantum in a seamless continuum is fundamentally redefining how we think about scientific discovery and innovation, allowing us to create more targeted solutions to drive R&D outcomes – Mark Roberts

How leaders can get ahead of the curve

Once leaders understand the full spectrum of value that comes as a result of the concurrence of AI and quantum, the next step is to look beyond incremental improvements and invest in integrated pipelines that leverage both technologies. We’ve outlined five strategic implications that executives must consider as they enhance their discovery pipelines:

1. Adopt realistic expectations: AI and quantum alone won’t eliminate the need for experimental validation or domain expertise. The technologies should be viewed as an E&D accelerator as opposed to an autonomous source of breakthrough discoveries.
2. Build hybrid teams: The most successful integrations of AI and quantum come from organizations that develop hybrid teams of data scientists and domain experts. Neither group can achieve desired results on their own, which present significant considerations for hiring, organizational structure, and knowledge management.
3. Use data as a strategic asset: High-quality, scientific data is a competitive advantage. Organizations with proprietary, well-structured datasets will outperform their competitors and should consider data strategy as a top priority.
4. Invest in computational infrastructure: First principles calculations require vast computing resources. Organizations must integrate high-performance computing capabilities into their pipelines to support AI and quantum mechanical modeling.
5. Adopt improved validation frameworks: Implement rigorous validation protocols for AI-generated discoveries. Multiple computational and experimental checks should be a standard practice to avoid pursuing false leads that drain resources.

The path to tomorrow

The next era of scientific discovery will be defined by integration. AI brings speed and scale, while quantum delivers depth and accuracy. Together, they create a discovery pipeline that is not just fast, but more dependable. As quantum computing capabilities grow, this synergy will only deepen, expanding what’s possible. In acting now by investing in people, data, and infrastructure, leaders can shape breakthroughs that will define the next generation of scientific and technological progress.