People Managing People

The episode delivers solid, substantive claims backed by data: the 35 vs. 700 application gap, the 81% executive vs. 9% worker trust discrepancy, 37% of workers skipping AI entirely, and 29% workflow accessibility rate. However, many insights are incremental observations about AI adoption rather than genuinely novel (e.g., 'traditional training doesn't work,' 'work is fragmented'). The pacing includes some repetition and throat-clearing that dilutes density.

“Executives think their employees are using around thirty-five applications. The actual number might surprise you...it's almost seven hundred applications”

“eighty-one percent of executives say that AI has improved productivity 'cause they see the outcome. But the workers were reporting a lot of AI-related friction. And only nine percent of workers actually trust AI for complex tasks”

The reframing of 'shadow AI' as 'business-led AI' is genuinely useful and counterintuitive. The emphasis on orchestration over model sophistication is contrarian to dominant market narrative. However, core arguments (AI needs context, implementation > capability, cognitive burden exists) are now fairly well-rehearsed in B2B tech discourse. The guest adds texture but limited truly fresh thinking.

“I actually don't like the term shadow AI. I prefer talking about...referring to it as business-led AI because that's what it is”

“The winners may not be the organizations with the most advanced AI, but the ones that best orchestrate people, systems, workflows, all that context”

Ofir Bloch is the head of a data-backed research organization (WalkMe) with five years of longitudinal survey data spanning hundreds of organizations. He speaks from systematic evidence, not anecdote. However, he is not a founder of a company using AI at operational scale, nor does he discuss personal implementation experience - he is primarily a vendor-adjacent analyst. Credible but not the highest tier of practitioner caliber.

“our report surfaces a lot of those points”

“This is the fifth report that we put out”

Strong use of named statistics and concrete numbers: 35 vs. 700 apps, 81% vs. 9% trust gap, 37% workers skip AI, 29% workflow accessibility, 1.9 - 3.7x confidence lift. The spot bonus example is vivid and concrete. However, the report lacks named company examples, specific dollar figures on AI spend (only 'fifty-four million dollars last year' stated vaguely), and timeline details on ROI failures. Some claims remain somewhat abstract (e.g., 'orchestration matters').

“Executives...would say a handful, right? You know, thirty, thirty-five applications. But in reality...it's almost seven hundred applications”

“Workers that had that in-flow kind of contextual support are one point nine to three point seven times more likely to report full confidence across every dimension of their work”

The host asks sharp follow-up questions ('how does leadership become so disconnected?' 'are you misdiagnosing the problem?') and does push back gently on framing. However, most follow-ups are soft and invite agreement rather than genuine disagreement or stress-testing. The host doesn't challenge Bloch on whether WalkMe's research has inherent bias (the company sells adoption software), nor does he press on alternative explanations for AI underperformance. The conversation is warm but lacks edge.

“I'm curious to know, at what point do we sort of stop treating this as a temporary adoption curve and maybe just admit that there's something fundamentally flawed in how we're implementing AI?”

“Why is context becoming the defining constraint in sort of enterprise AI?”