> ## Content Index
> Fetch the complete content index at: https://the-human-capitalist.ghost.io/llms.txt
> Use this file to discover other available public pages before exploring further.

# Is AI the Threat to HR, or Our Refusal to Change?
- URL: https://the-human-capitalist.ghost.io/is-ai-the-threat-to-hr-or-our-refusal-to-change/
- Published: 2026-08-23T23:00:00.000Z
- Updated: 2026-08-25T01:18:13.000Z
- Description: A depth coach says AI is not HR's biggest risk. Trent Cotton and Tony Dina break down deferred judgment and the Refusal Gap for C-suite leaders.
- Author: Trent Cotton
- Tags: Podcast Recaps

## **TL;DR**

**Short answer:** It depends on what HR chooses to do with its own judgment, and the bigger risk right now is refusal rather than replacement. Tony Dina, a former corporate technology leader with nearly 30 years in high tech who now works as a Jungian depth coach, argues that the danger shows up the moment we accept the first paragraph a model produces as correct and quietly hand over the judgment that made us valuable in the first place.

**Key stats you need to know:**

- The average manager's span of control climbed from 10.9 direct reports in 2024 to 12.1 in 2025, up from 8.2 in 2013, which is roughly a 50% increase in a decade ([**Business Insider**](https://www.businessinsider.com/middle-managers-have-more-direct-reports-after-great-flattening-2026-1?ref=the-human-capitalist.ghost.io)).
- 56% of workers say their employer has not consulted them about how AI tools get used in their work, and optimism about AI's impact on workers dropped 10 percentage points in a single year ([**Jobs for the Future**](https://www.jff.org/newsroom/press-releases/worker-anxiety-over-ai-is-growing-and-employers-arent-preparing-employees-for-whats-next-new-survey-finds/?ref=the-human-capitalist.ghost.io)).
- Roughly 70% of enterprise transformations fail, and 72% of companies with failed programs point to employee resistance or management behavior as the primary barrier ([**McKinsey**](https://www.mckinsey.com/capabilities/people-and-organizational-performance/our-insights/the-organization-blog/organizations-do-not-change-people-change?ref=the-human-capitalist.ghost.io)).

**The leadership takeaway:** Over the next 12 to 24 months, the CHROs who win will be the ones who protect human judgment as deliberately as their CIO protects uptime, because efficiency is already assumed and judgment is what your board will actually be short on.

---

> 📺 Watch the full episode on YouTube: [**https://youtu.be/ZDXbU8qK1q4**](https://youtu.be/ZDXbU8qK1q4?ref=the-human-capitalist.ghost.io)

> 📻 Listen to it on Spotify: [**https://open.spotify.com/show/57xtYAUsIYrg7gQ0a6RFwc**](https://open.spotify.com/show/57xtYAUsIYrg7gQ0a6RFwc?ref=the-human-capitalist.ghost.io)

---

### **Is AI the Threat to HR, or Is Refusal the Real Threat?**

Refusal is the threat, because AI only removes what we willingly stop doing. I brought this to [**Tony Dina**](https://www.linkedin.com/in/antdina?ref=the-human-capitalist.ghost.io), a trained painter with an MBA who spent nearly three decades leading data and analytics teams in high tech before becoming a Jungian depth coach at his practice Crucial Canvas. He calls the risk deferred judgment. Cognitive offload has a long and healthy history, from the grocery list to the calculator, and he told me plainly that HR should embrace it. The trap opens when we assume the first output is right and stop checking our own thinking.

The research is catching up to that instinct. MIT Media Lab researchers tracking essay writers over four months found that participants relying on a large language model consistently underperformed at neural, linguistic, and behavioral levels compared with those working unassisted, a pattern the team labeled cognitive debt ([**MIT Media Lab**](https://www.media.mit.edu/publications/your-brain-on-chatgpt/?ref=the-human-capitalist.ghost.io)). This was the second conversation I had that same day about deferred judgment, which told me it has moved past coincidence and into pattern.

What makes this an HR problem is where the pressure lands. Tony described senior technology leaders convinced that some 25 year old with a stack of AI tools will build a billion dollar competitor without hiring anyone, so they are racing to push AI into every business process. That fear travels downhill fast, and it lands on managers already carrying more people than they did two years ago. I have written about the same mechanism in [**why AI adoption in HR stalls on skills, privacy, and resistance**](https://www.linkedin.com/pulse/why-ai-adoption-hr-stallsskills-privacy-resistance-trent-cotton-blfke?ref=the-human-capitalist.ghost.io). The blocker is rarely the tool.

The macro data reinforces it. The World Economic Forum reports that 39% of the skills employers require will change by 2030, with analytical thinking still the most sought after core skill and creative thinking climbing fast ([**World Economic Forum**](https://www.weforum.org/publications/the-future-of-jobs-report-2025/digest/?ref=the-human-capitalist.ghost.io)). Those are the capabilities that atrophy when judgment gets outsourced.

### **What Is the Refusal Gap, and Why Does It Matter in the Boardroom?**

The Refusal Gap is my term for the distance between the AI mandate an organization announces and the personal change its leaders are actually willing to make, and it is where most transformation quietly dies. I built it out of this conversation with Tony, who works from Joseph Campbell's hero's journey and points out that the first phase after the call to adventure is the refusal of the call. Three groups live inside that gap.

- **The Racers.** Senior leaders sprinting to AI-enable manufacturing, HR, marketing, and sales because they are afraid of a competitor that does not exist yet. They change the org chart and leave their own operating habits untouched.
- **The Squeezed.** Mid-tier managers absorbing a mandate from above and fear from below while their span of control widens from one to five, to one to ten, to one to thirty. Tony's warning was blunt. Push far enough and a knowledge-work leader becomes a people processor.
- **The Deferrers.** Individual contributors and HR practitioners who accept the model's first answer because it reads well, then lose the muscle that made their read of a situation worth paying for.

This belongs in the boardroom because every item carries a number. Span of control compression shows up in attrition and manager burnout. Deferred judgment shows up in quality escapes and bad hiring decisions no one can trace. Refusal at the top shows up as the 70% transformation failure rate, with 72% of failures traced to employee resistance and management behavior ([**McKinsey**](https://www.mckinsey.com/capabilities/people-and-organizational-performance/our-insights/the-organization-blog/organizations-do-not-change-people-change?ref=the-human-capitalist.ghost.io)). Tony's ampersand mindset is the honest version of the story. AI is a productivity gain and people are terrified. Both are true at once, and pretending otherwise breaks trust. In a boardroom, I would say it this way. Your AI plan is only as good as the leadership behavior you are willing to change alongside it, and right now that ratio is upside down. Same conclusion I reached in [**the CHRO playbook for closing the AI readiness gap**](https://www.trentcotton.com/blog/ai-readiness-gap?ref=the-human-capitalist.ghost.io).

### **How Should C-Suite Leaders Respond to the Refusal Gap?**

C-suite leaders and CHROs should respond by treating human judgment as a governed asset and by changing their own behavior before they ask anyone else to change theirs. Here is the playbook I would run out of this conversation.

1. **Change yourself before you change the operating model.** Tony's first rule was direct. If you want to change your business, change yourself, full stop. Write down the two decisions you personally made differently this quarter because of AI, then put that in front of your executive team before the next all hands.
2. **Name what AI cannot do, then staff for it.** Tony's second act covers creativity, human judgment, and connection. Audit your top 20 roles and tag which tasks require those three, then protect that time on the calendar the way you protect revenue reviews. When the market pays a visible premium for complex judgment, flat comp bands guarantee attrition.
3. **Use the four-season change message, in order.** Tony walks leaders through a sequence he illustrates with the changing seasons. Start with what is genuinely good about the current state and commit to keeping it. Get the team to admit the downsides of standing still. Introduce the new state and go straight to the risks, because that is what people already feel. Only then talk about the upside. Skipping the hard middle makes an announcement land as spin.
4. **Consult your workforce before you deploy.** With 56% of workers saying their employer never asked how AI should be used in their work ([**Jobs for the Future**](https://www.jff.org/newsroom/press-releases/worker-anxiety-over-ai-is-growing-and-employers-arent-preparing-employees-for-whats-next-new-survey-finds/?ref=the-human-capitalist.ghost.io)), the cheapest engagement lever available is a structured conversation with the people who will use the tools. Same case I made in [**what is really behind the wall of resistance to AI in HR**](https://www.linkedin.com/pulse/great-wall-resistance-lets-get-real-barriers-ai-adoption-trent-cotton-ftooe?ref=the-human-capitalist.ghost.io).
5. **Move from goal driven to gift driven.** Tony's line stuck with me. We should be gift driven rather than goal driven, which puts the leader's job at finding where someone's real capability meets a real problem. Drop the over-scheduled goal stack, hold the vision, and let people opt in. That is the retention play most executives leave on the table while they redesign the org chart around project-based work, which I covered in [**what the Hollywood Model means for HR leaders**](https://www.trentcotton.com/blog/hollywood-model-of-work?ref=the-human-capitalist.ghost.io).

### **What This Means for You as a Leader**

The tension underneath this episode is that AI rewards efficiency while quietly exposing how little judgment we have left in the system. Tony put it in terms I will not forget. A young psychologist once told him that if you do not take care of your demons, they go into the basement of your soul and start lifting weights, and he sees a lot of leaders with their backs pressed against that door. Organizations behave the same way. Unaddressed fear does not disappear. It gets stronger while everyone talks about productivity gains.

Winning here looks unglamorous. Leaders who close the Refusal Gap keep their spans of control humane, keep human judgment in the loop on decisions that carry consequence, and tell their people the truth about the hard middle before promising the upside. Leaders who skip that work will hit their efficiency targets and then wonder why the innovation stopped and the good people left. I would rather be uncomfortable now than confused later, which is what the credibility work in [**rebuilding the CHRO seat after Bolt fired its HR team**](https://www.linkedin.com/pulse/bolt-fired-its-hr-team-how-chros-survive-trent-cotton-awmce?ref=the-human-capitalist.ghost.io) comes down to.

I am still thinking about his candle and train explanation of non-dualism at two in the morning, which is a good sign for a podcast conversation. The full episode covers the four archetypes, the three acts of what AI cannot do, and why the monster in the hero's journey is never on the outside. Catch it on [**this week's episode of The Human Capitalist podcast**](https://youtu.be/ZDXbU8qK1q4?ref=the-human-capitalist.ghost.io).

### **FAQ**

### **Is AI destroying HR jobs or creating them?**

Both, and the split depends on judgment. The World Economic Forum projects that 39% of required skills will shift by 2030 with analytical and creative thinking leading demand ([**World Economic Forum**](https://www.weforum.org/publications/the-future-of-jobs-report-2025/digest/?ref=the-human-capitalist.ghost.io)), so the roles that survive are the ones where human judgment stays in the loop.

### **What is deferred judgment in AI adoption?**

Deferred judgment is Tony Dina's term for the moment a professional accepts a model's first output as correct without evaluating it. MIT Media Lab researchers found that people relying on a language model over four months underperformed unassisted peers at neural, linguistic, and behavioral levels ([**MIT Media Lab**](https://www.media.mit.edu/publications/your-brain-on-chatgpt/?ref=the-human-capitalist.ghost.io)).

### **Why are middle managers under so much pressure right now?**

Because they are absorbing a mandate from above and fear from below while their teams grow. Average span of control rose from 10.9 reports in 2024 to 12.1 in 2025, against 8.2 in 2013 ([**Business Insider**](https://www.businessinsider.com/middle-managers-have-more-direct-reports-after-great-flattening-2026-1?ref=the-human-capitalist.ghost.io)).

### **What is the Refusal Gap in simple terms?**

The Refusal Gap is my framework for the distance between the AI change a company announces and the personal change its leaders will actually make. It shows up in three groups: the Racers at the top, the Squeezed in the middle, and the Deferrers who hand their judgment to the model.

### **How does closing the Refusal Gap improve job quality?**

It replaces vague reassurance with an honest sequence. Naming what is good about the present, admitting the downsides of standing still, and stating the risks of the new state before promising the upside gives people something they can trust, which matters when 56% say they were never consulted about AI in their own work ([**Jobs for the Future**](https://www.jff.org/newsroom/press-releases/worker-anxiety-over-ai-is-growing-and-employers-arent-preparing-employees-for-whats-next-new-survey-finds/?ref=the-human-capitalist.ghost.io)).

### **What should CHROs and CEOs do first?**

Audit which decisions in your top 20 roles still require human judgment, then protect that time explicitly. Do that before any new AI rollout, and pair it with a documented account of what you personally changed as a leader this quarter.