The SafetyPro Podcast

The episode surfaces several genuinely useful practitioner concepts - algorithmic overconfidence risk, the 'Yelp effect' as a data-completeness metaphor, de-skilling in high-automation environments, and the psychosocial harm of surveillance culture - but these are buried in heavy rambling, self-interruption, and lengthy filler that dramatically reduce the ideas-per-minute ratio.

“when workers and managers start deferring to the algorithm instead of exercising their own professional judgment, then you transfer risk from the hazard to the system”

“if our near miss reporting rate is low. The AI is learning from incomplete data or getting an incomplete picture”

The 'AI is a tool not a replacement' frame is well-worn, and most of the AI cautions (bias in training data, privacy concerns with wearables) are widely circulated; however, the observation that QR-code reporting programs can collide with existing no-cell-phone policies, and the pedometer-gaming anecdote as a concrete analogy for surveillance backfire, are fresher and grounded in practice.

“How many of y'all have policies in your organizations that prohibit the use of cell phones in the production area? Right? Probably all of you.”

“they're motivated by just simply trying not to trick ⁓ trip a flag or on the algorithm or on the monitoring”

This is a solo monologue by a practitioner-host with apparent consulting and in-house safety experience; there are no guests to evaluate, which caps the ceiling, and the host's seniority and scale of experience remain undemonstrated beyond anecdotal references to past clients.

“I worked with a company where ⁓ they wanted to improve and it was it was a really good workflow”

“I know a a company that worked that piloted a lot of ⁓ VR training”

There are some concrete data points (20% VR motion sickness industry figure, 80% root causes attributed to worker error at one client, $100k quoted for manual LOTO conversion) and a citation of a named researcher and NSC report, but the citations are fumbled mid-sentence, the researcher's name is mangled, and the NSC '2026' report date is oddly forward-dated, undermining credibility.

“The National Safety Council, in their 2026 report on emerging workplace safety technologies, highlights computer vision and predictive risk modeling”

“about 20% of the population, like this will make them ill”

As a solo episode there is no host-guest dynamic, no follow-up questions, and no productive disagreement; the host repeatedly loses his train of thought, cannot recall citations or company names mid-sentence, and the structure meanders without sharp transitions, leaving the listener to do significant work to extract the substance.

“Pul Pilati, Pilati and colleagues, Pilates and colleagues. I've got multiple monitors here, so I'm gonna try to bring everything over to my my left side”

“I don't even know what my own episodes are called”