Math plus AI -- a way to control agents
Most enterprises are familiar with probabilistic AI systems such as generative AI and agentic AI. According to Maha Achour, CEO and founder of enterprise AI platform vendor Kodamai, another form of AI could help enterprises better trust these systems: math-based AI. In this episode of Targeting AI, Achour discusses how math-based AI can be trusted more than other forms of AI because it is harder to hack or be manipulated. She dives deep into the mathematical principles that Kodamai uses, including category theory and type theory. In this episode, we discuss: Mathematical principles such as category theory and type theory could help remove the black box behind current forms of AI. The importance of using neuro-symbolic AI. What grounding AI systems in mathematical certainty rather than probabilistic approximations means for challenges such as hallucinations, governance and security. Why artificial general intelligence and artificial superintelligence require human collaboration. The superiority of human intuition. To learn more about generative and agentic AI, check out AI Business from Informa TechTarget, and please subscribe to our newsletter to keep up to date on the most important AI news. To watch video clips from our podcast, subscribe to our YouTube channel, @EyeonTech . References: Startup Pioneering Neuro-Symbolic AI Secures Bridge Funding Mathematical Superintelligence Startup Valued at $1.45B An Explanation of the Different Types of AI


