Systems That Set You Free: Interviews with Leaders in Consulting

The episode lands one genuinely useful operational insight - that AI tools built on SOPs rather than observed practice will produce 'half correct output every single time' - but spends most of its runtime restating the same point (talk to your team) in slightly different ways, with a promotional pitch eating the final third.

“there is a gap between your SOP and what your team actually does. And if you're building your AI on the SOP without consulting the team, you're going to get a tool that produces a half correct output every single time”

“quiet is where sabotage lives”

Reframing AI rollout failure as a leadership/people problem rather than a technology problem is a modestly contrarian and useful angle, but the underlying framework (before/during/after conversations, change management, feedback loops) is standard organisational change advice dressed in AI vocabulary.

“That is not an employee problem. That is a you problem. And that is a leadership problem”

“Not to sell them on it, not to announce it, to ask them”

This is a solo monologue with no guest; the host claims 20 years across Booz Allen Hamilton and Deloitte in 30 countries, which are real credentials, but the episode functions largely as a marketing vehicle for her own diagnostic service, undermining the credibility of the practitioner framing.

“I'm Latonya Roberts. I'm a fractional chief operations officer and hybrid workforce architect. And for over 20 years, I have been inside organizations from Booz Allen Hamilton to Deloitte, across 30 different countries”

“That is exactly what I look at. In the Operational Freedom Diagnostic”

One thin anonymised case study (5-person consulting firm, 2 hours saved per report) provides some concrete grounding, but the headline statistic ('30% of employees are actively sabotaging AI rollout') is presented with zero source attribution, and no company names, tool names, or financial figures appear anywhere else.

“30% of employees right now in companies across the US, the UK and across Europe are actively sabotaging their company's AI rollout”

“The team saved two hours per report, per client, every single time”

As a solo monologue there is structurally no host-guest dynamic, no follow-up questioning, and no pushback; the rhetorical questions posed ('Is this tool bad? Did I waste money?') are used as setup devices rather than genuine inquiry, and the episode closes as a sales pitch.

“She's asking the wrong questions”

“So the GPT was built on incomplete information and it produced incomplete output. Garbage in, garbage out. Not that it was garbage, but you get the point”