Episode Details

We welcomed guest Bryant Cruz, founder of New Sapience, a company developing artificial intelligence technology that processes language in a fundamentally different way from traditional models such as ChatGPT and GROK. New Sapience’s approach involves building internal models of the world that mirror human knowledge, allowing their AI to reason and understand more like a person rather than relying on statistical language patterns.

To start the interview, I offered Bryant a quick summary of the key difference in New Sapience’s technology compared to a traditional AI company in that New Sapience does not rely on statistical word associations like current mainstream AI models. Instead, as then explained by Bryant, their system creates an internal, structured representation of knowledge—more akin to how humans comprehend and retrieve information. He emphasized that while models like ChatGPT often produce plausible responses, their accuracy can be misleading, and users should be cautious when relying on them for critical tasks. Read the full summary of this program at www.thespaceshow.com for this program on this date, July 11, 2025. You can also read the summary on our Substack page, doctorspace.substack.com.

Bryant shared his journey into AI, shaped by his experiences as a naval aviator and an engineer on the Hubble Space Telescope. In the mid-1980s, he developed an expert system to streamline spacecraft operations, inspired by the efficiency of aircraft systems. This innovation utilized tabular displays and real-time data but was ultimately limited by scalability challenges. Realizing the need to embed knowledge rather than just data processing into machines, he founded a new company to explore advanced AI applications for space systems.

In the 1990s, Bryant developed a state-based markup language for spacecraft control systems. This tool allowed engineers to describe spacecraft operations as state machines, significantly reducing costs related to launch integration and testing. Despite its technical advantages, the system wasn’t adopted by government or military institutions due to a preference for maintaining large control room staffing. Compounding the challenge, the market for satellite communication was disrupted by the expansion of undersea fiber-optic cables which were far more cost effective. Bryant eventually sold this technology.

Bryant reflected on the evolution of AI, including the decision to avoid the term "AI" early on due to industry disillusionment. His team instead focused on automation technology and the curation of computable human knowledge, which became the foundation for New Sapience. While the company has yet to secure clients in the space industry, its platform is now ready for real-world deployment, offering a common-sense, language-independent knowledge system.

I presented a listener email question asking about New Sapience’s relevance to self-driving cars, such as those operated by Waymo. Bryant acknowledged the strengths of current vehicle AI in detecting and reacting to programmed scenarios, but stressed their limitations in unfamiliar situations. He compared the ideal AI to a “higher brain function,” much like a cowboy guiding cattle. He also discussed the compact nature of their knowledge structures, making them suitable for autonomous spacecraft.

Bryant introduced Sapien, the core technology aiming to endow machines with human-like understanding. He highlighted its ability to store and retrieve curated knowledge efficiently—something traditional education systems or data-heavy AI cannot match. Applications for Sapien include telepresence, remote operations, and a forthcoming consumer product for personal assistance and memory support. A caller, Fremont John, engaged Bryant in a thoughtful discussion about