Stepsero

The episode delivers core insights about psychological safety and AI adoption, including the hospital team study and the distinction between 'what' vs. 'who' questions for mistake reporting. However, it relies heavily on platitudes in the second half ("be transparent," "reskill," "reinvent") and the three principles for AI-era psychological safety (experiment, curiosity, humanity) are stated without deep exploration or novel angles that a B2B operator wouldn't already recognize.

“Psychological safety is a work environment where employees feel safe to express their questions, concerns, ideas and mistakes.”

“make it safe to do experiments, make it safe to show curiosity, and make it safe to show what makes us humane.”

The core frameworks are well-worn: Edmondson's definition, the hospital study (decades old), and the Google dogfooding example are standard references in psychological safety discourse. The attempt to connect psychological safety to AI is topical but superficial - the three principles (experiment, curiosity, humanity) lack counterintuitive depth or contrarian positioning. No fresh first-principles thinking emerges.

“Psychological safety is a work environment where employees feel safe to express their questions, concerns, ideas and mistakes. It's a key aspect of high performing teams”

“There's a very famous research in 1990s, researchers are looking at two teams in a hospital”

Mehmet Baha is positioned as a global trainer, speaker, and author with consulting experience, but the transcript reveals limited operational depth or evidence of having built or led organizations through large-scale psychological safety or AI transformation initiatives. He speaks from a training/advisory perspective rather than as a practitioner who has implemented these principles at scale in high-stakes contexts.

“Mehmet Baha, uh, global trainer and speaker and the author of Creating Psychological Safety at Work”

“You consult on them, you speak a lot about both of them”

The episode cites Gallup statistics (27% turnover reduction, 40% safety incident reduction, 12% productivity increase) and references the hospital study, Google's dogfooding, and a ChatGPT book by a first employee. Beyond these, claims remain abstract: no named organizations implementing the AI + safety framework, no timelines, no cost figures, no detailed case studies of the transition struggle leaders face. The advice is largely un-anchored to concrete organizational examples.

“27% reduction in employee turnover, 40% reduction in safety incidents, 12% increase in employee productivity”

“One example is Google. Google has a tool called Dog Fooding.”

The host asks thoughtful bridging questions (psychological safety + AI, leadership blockers, employee silence as performance risk) and references the guest's book to build credibility. However, follow-ups are generally soft. When the guest gives vague answers (e.g., "I'm not sure if it's difficult or more easy") or pivots with platitudes, the host doesn't press for specificity or challenge underlying assumptions. The conversation feels more like a friendly promotional interview than rigorous exploration.

“And we know that AI is generating a lot of stress and anxiety right now. There is this fear of being replaced, there is this pressure to upskill very fast. Do you think that it is harder to build a psychological safe organization in this day and age because of what is happening with AI?”

“Do you see any specific blocker or pushback or obstacle that prevents organizations from showing safety around experimentation, curiosity and humanity?”