Digital Value Creation

The episode contains several substantive ideas about hybrid AI, the explainability problem in enterprise systems, and realistic AI adoption patterns, but much of the discussion is exploratory and conversational rather than densely packed with novel claims. The McKinsey survey findings and the framework of deterministic vs. probabilistic AI add concrete substance, but there is also considerable throat-clearing and meandering between topics that dilutes insight-per-minute.

“We're building the unexplainable enterprise... unexplainable quantum computing and unexplainable generative models”

“What we have seen that some of the early coding models really did not resonate with our top Developers because they created code snippet, there was some improvement in performance, not a true breakthrough”

The hybrid AI framework (deterministic vs. probabilistic processes) is a useful lens, and the observation about tech industry shipping solutions without defined problems is sound but not novel. The comparison to pharma/chemical industry practices is mildly contrarian. However, much of the AI agent discussion and concerns about explainability are well-trodden ground in 2025; the episode recycles common venture hype cycle critiques and doesn't present genuinely counterintuitive or first-principles arguments.

“In the tech industry we do this all the time. So instead of us going out and solving problems, which is what a pharmaceutical industry would do”

“We are amazing at creating solutions without defining the use case for them. And it's like well it can solve anything and everything”

Speaker B works at an AI-focused hardware company and has attended CIO roundtables and AI forums, suggesting some relevant enterprise exposure. Speaker A has worked at an RPA company and recently attended multiple summits, indicating practical industry engagement. However, neither is named or positioned as a recognized operator or executive at a major company; they are framed as internal speakers at their own firms rather than proven practitioners at scale. The lack of specificity about their seniority and actual shipping-at-scale experience limits the caliber rating.

“My brother works at an AI focused hardware company and I work at an AI focused software company”

“I just attended a couple CIO roundtables and AI forums”

The episode cites the McKinsey Quantum Black survey (1500 companies, March 2025) with concrete percentages: 55% of tech companies using GenAI in sales/marketing, 30% in cogeneration, 40% of consultants using AI for knowledge management. The banking loan-processing example and references to specific tools (Anthropic Claude 3.7, Gemini, Klein, Perplexity) provide some grounding. However, many claims about company adoption, conference discussions, and customer insights lack naming or hard numbers; the episode relies heavily on secondhand reporting from "summits" and "forums" without attribution.

“In the technology industry itself is 55% of the companies saying we are using generative AI in sales and marketing”

“cogeneration, uh, which is very common is about 30% of the companies said we're using that in the tech industry”

The conversation is collegial and covers multiple angles (explainability, regulation, career impact, adoption patterns), but lacks sharp follow-up questions or productive friction. Speaker A often asks open-ended softballs ("how do you think about that?", "what do you think it means?") without pushing back on claims or drilling into inconsistencies. When contradictions surface - e.g., why top developers didn't adopt early AI tools - Speaker B explains it away smoothly without real challenge. The hosts do not rigorously test each other's assertions.

“So that's what I'm finding. Um, uh, and there was an interesting thing as it relates to all of this, uh, because I was in Europe”

“How do you think about those two ends of the, of the spectrum”