Human Capital Leadership

The episode develops a central thesis about judgment formation being undermined by AI automation of early-career work, supported by concrete industry examples (audit firms, medical residency, flight training). However, substantial portions are devoted to scene-setting, book promotion, and affirmations rather than densely packed novel claims. The core insight about the 'judgment void' and the four dimensions of human work (declarative knowledge, procedural tasks, judgment, relationships) is valuable but expanded at length without adding proportional depth.

“when there's ambiguity, make a decision...that's the dismantling of the one thing it can't replace”

“40% of the work that our auditors do next year will be done by AI”

The framing of judgment as the irreplaceable human capability under threat from AI automation is reasonably fresh, as is the structural analysis of organizational pyramids flattening into diamonds. However, the core argument about AI handling routine work while humans must handle ambiguity, relationships, and wisdom is increasingly common in business discourse. The episode lacks contrarian thinking or first-principles challenges to prevailing AI narratives.

“AI is so compelling...it's a probabilistic model”

“medical...residency program...flight training”

Larry Durham is a legitimate practitioner with 35 years in learning and talent development, 10 years at PwC, and direct consulting relationships with Big Four firms and Fortune 500 companies. He brings real operational experience rather than pure theory. However, he is primarily a consultant and author rather than a line executive or operator who directly built business units or managed P&Ls at scale.

“Over 30 years he has partnered with large professional services firms and Fortune 500 companies”

“we work across many industries...we do spend a lot of time...consulting and tax and audit”

The episode provides concrete industry examples (audit firms doing 40% AI work, Big Four clients, residency and flight training programs) and specific methodological claims (90+ defined judgments in audit, simulations with edge cases and bias testing). However, most evidence is anecdotal rather than quantified. Missing are: specific company names beyond 'Big Four', data on judgment-formation success rates, timelines for the regulatory requirements mentioned, or metrics on the diamond model's viability.

“40% of the work that our auditors do next year will be done by AI”

“unemployment rate amongst early career...three years out of college is double what the rest of the population is”

The host asks reasonable opening questions and provides appropriate setup, but rarely pushes back on claims or probes deeper. When Durham makes assertions (e.g., about higher ed's role, the viability of simulations, the diamond model's failure timeline), the host largely affirms and moves forward rather than challenging specifics. The conversation reads more as a book-promotion format with soft follow-ups than as genuine adversarial inquiry.

“Yeah, interesting. And the relationship piece, the, the wisdom, the, the ability to deal with ambiguity”

“I love it. I love it Larry.”