Medium Article: https://medium.com/@jsmith0475/what-holds-an-ai-together-063fcb26c876

"What Holds an AI Together? The case for vertical causality in machine intelligence," by Dr. Jerry A. Smith, argues that contemporary artificial intelligence systems are fundamentally incomplete because they rely solely on horizontal causality, which governs the sequential flow of actions and feedback across time. This reliance on the temporal axis results in systems that are locally competent but lack global coherence, leading the author to introduce the concept of vertical causality. Vertical causality describes simultaneous, structural dependencies—such as the underlying architecture, goal representations, and identity models—that sustain the system and ground its purpose at every moment action occurs. The author explains that achieving genuine artificial agency requires integrating both dimensions in a "duplex ecosystem," where vertical structures define the space of possible behaviors while horizontal processes explore it. Consequently, robust AI alignment should focus not just on sequential checks but on the architecture itself, ensuring that essential commitments are structurally operative rather than merely procedural outcomes.