Zapata AI touts generative AI applicability as trading debut edges closer
Zapata AI, a company that started life in 2017 with a heavy focus on quantum computing before refocusing on the intersection of quantum and generative AI, is a step closer to becoming a publicly traded firm after shareholders of Andretti Acquisition Corp. last week approved Andretti’s previously announced merger with Zapata AI.
Andretti is a publicly traded special purpose acquisition company (SPAC), and the approval of the merger, which also already has earned SEC approval, clears the path for Zapata to debut on the stock market. How soon that will happen remains unclear for now, as Andretti has not yet announced a debut date for the combined company’s stock, but it would make Boston-based Zapata AI at least the sixth firm with a strong quantum background to hit the stock market in the last three years. Four of those–IonQ, Rigetti Computing, D-Wave, and Quantum Computing Inc., fit the description of “pure-play” quantum companies, while Zapata last year pivoted from a broad quantum computing focus to the more specific mission of applying quantum algorithms to industrial-scale generative AI problems.
The Andretti-Zapata merger was announced in September 2023, and followed a previous partnership that Andretti Autosport and Zapata had commenced in early 2022 to explore use cases for quantum computing. While much of Andretti’s early and ongoing interest in Zapata AI has to do with how its technology could be applied to computations designed to improve auto racing performance and efficiency, the merger partners have acknowledged that the see a large addressable market that includes generative AI computations and problems across multiple industries.
During a recent webinar for investors and analysts, Zapata AI CEO Christopher Savoie frequently used the phrase “quantum math” to describe what his company brings to the generative AI table.
“Our computational approaches leverage the statistical advantages of quantum math and other techniques,” Savoie said, later adding, “Our technology is derived from math inspired by quantum physics… The hard part is turning that discipline of physics into useful technology. Fortunately, our work in this area has many transferable and positive implications for generative AI. Being experts at quantum math.. allows us to enhance what we believe are the key desirable qualities of generative models. Namely, quantum statistics can enhance generative models’ ability to generalize or extrapolate missing information, and generate new high-quality information, as well as generate a greater range of solutions [to problems]. This is called ‘expressibility.”
Savoie noted projects on which Zapata AI has worked with partners to demonstrate these benefits, including research published last June with Insilico Medicine, Foxconn, and the University of Toronto that was based on a project that explored the use of hybrid quantum-classical generative adversarial networks (GAN) for drug discovery applications. “Our research showed that generative models enhanced with quantum components generated more desirable drug-like molecules than those generated by a traditional generative model,” Savoie said.
He also said that in addition to that project and others that have used Zapata AI’s software running on classical computers, the company also has run demonstrations on quantum computers, including a past project with IonQ and another recently announced in partnership with D-Wave. Savoie suggested that as quantum computers become more mature and commercially viable, that evolution will translate into an even bigger payoff for companies like Zapata AI as they apply quantum techniques to generative AI and machine learning.
“The promise of quantum computers is that we’re going to be able to do linear algebra–matrix multiplication, if you will–and it will be much faster and much more accurate on large dimensional spaces once we have quantum computers that are large enough and accurate enough to do that faster than classical computers,” Savoie said, noting the companies like Microsoft also have been talking about how quantum will benefit AI efforts. “Using quantum math is better. It’s slower on GPUs today, but it’s still faster than the alternatives, and it will be even better and faster… We will be able to do even more and get even better answers–more accurate answers and faster answers–when we have quantum computers actually working.”
Zapata AI began in 2017 as Zapata Computing, a spin-off from a Harvard University lab, and though it mentioned AI and machine learning from the outset, it was more often identified as a quantum computing company until it formally changed its name to Zapata AI sometime during 2023.
Image: Screen capture from Feb. 8 Zapata AI webinar for analysts and investors. Pictured are Zapata CEO Christopher Savoie, Zapata VP of Marketing and Communications Mick Emmett, and Andretti Acquisition Corp. Co-CEO Bill Sandbrook.
Dan O’Shea has covered telecommunications and related topics including semiconductors, sensors, retail systems, digital payments and quantum computing/technology for over 25 years.
