In this episode, we sit down with John DeMartino, a Staff Solutions Engineer at Google Cloud working on the incubation of next-generation AI and large language model products.
The conversation cuts through AI hype to explore what AI really is, what it isn’t, and why adoption breaks down in enterprise and regulated environments like accounting and finance. We discuss how consumer tools reshaped expectations, the risks of nondeterministic systems, and why explainability, guardrails, and workflow design matter more than model buzzwords.
John shares practical insight on blending algorithms, automation, and AI responsibly, the dangers of “impostor AI,” and how rushed implementations can stall progress for entire industries. We also examine the structural pressures facing accounting firms today and how AI can relieve capacity constraints without introducing new risk.
This episode offers a clear, realistic view of where AI is headed and what organizations need to understand to stay relevant over the next several years.