Skip to main content

Boosting AI with Quantum Technology

The Generative AI era has stressed the limitations of classic computer architecture, leading researchers to embrace quantum computing. Quantum improves generative AI’s performance and unlocks new capabilities unobtainable without the marriage of the two technologies. Combining the two can solve complex business problems, enrich business analytics with accurate and inferred data, and customize large language models for customer-specific problems, making them orders of magnitude more efficient.

“It’s quite interesting to show the opportunities of quantum and generative AI for different camps of applications,” said Marta Mauri, Research Manager, Quantum AI, Zapata AI. “We did work with BASF, the largest chemical company, to utilize generative AI and machine learning for materials discovery. We had projects with BMW and a food and beverage company where the focus was optimization of scheduling, with manufacturing plant scheduling in the case of BMW and delivery operations for the food and beverage company.”

Other clients Zapata AI has worked with include Andretti, European financial institution BBVA, BP, and DARPA. Complex optimization problems benefit from the application of quantum computing while generative AI provides the ability to generate new data and ways to solve problems.

“It’s not a fight between classical and quantum,” said Mauri. “It’s a synergy between the two, connecting two worlds. We can enhance classical with quantum and these things work really well together.”

Zapata has conducted research in the best ways to evaluate and leverage the combination of generation AI and quantum computing, efforts that are especially important given the scarcity and limitations of current quantum hardware and the classic hardware intensive nature of generative AI. The company has been able to use quantum plus generative AI models to more effectively evaluate chemical combinations for new drugs.  

“Drug development is quite complicated in the sense that it is time consuming, resource consuming, and it has a very low probability of success,” said Mauri. “You have to come up with new molecules, synthesize them, test them. If you have to put something together and test it, it’s very expensive in terms of resources and time. It usually takes 10 years for a new medicine that can be created from initial discoveries to the marketplace.”  

Quantum computing can generate viable compounds quicker, providing a boost to generative AI engines searching for new solutions for medicines, battery improvements, and other material challenges that were traditionally conducted on a trial-and-error basis in the laboratory. For more information on the ways quantum computing can boost generative AI, listen to the latest Fiber for Breakfast podcast.