When the executive team at PG&E stared down a bankruptcy proceeding while simultaneously navigating devastating wildfires and unprecedented liability exposure, no algorithm could have charted the path forward. That required something AI will never possess—battle-tested human judgment forged in the crucible of crisis.
I learned this truth while managing talent strategy through catastrophic wildfires and bankruptcy at PG&E, steering recruiting operations during COVID-19’s unprecedented disruption at US Foods, and integrating cultures during M&A transitions at Kerry Group, CHC Helicopter, and Sara Lee. After two decades in Fortune 150 environments and now helping mid-market companies transform their recruiting functions, I’ve witnessed a fundamental reality: AI is a powerful accelerant, but experience is the irreplaceable foundation.
When California’s Camp Fire devastated Paradise on November 8, 2018, it became the deadliest wildfire in California history. The scale was staggering: 85 people killed, 18,804 structures destroyed, 153,336 acres burned, and more than 52,000 people displaced. On that first day, 95% of Paradise had been consumed by flames moving at 80 acres per minute.
PG&E’s fatigued transmission equipment had sparked the inferno. Within weeks, the company faced $30 billion in estimated liabilities from this fire and the 2017 Wine Country fires combined. On January 29, 2019, we filed for Chapter 11 bankruptcy—the company’s second bankruptcy in two decades, but this time under circumstances that threatened our very existence.
For our talent acquisition team managing a 24,000-person workforce responsible for keeping electricity flowing to 16 million Californians, this wasn’t a scenario you could model with predictive analytics. The decisions we made weren’t about optimization algorithms or retention risk scores. They were about reading fear in people’s eyes during town halls, understanding which leaders would hold steady when media helicopters circled overhead and protesters gathered at our headquarters, and knowing how to communicate hope without dishonesty.
We weren’t making hiring decisions based on growth forecasts. We were triaging which roles were mission-critical to keep the lights on during a multi-year Chapter 11 proceeding while facing criminal investigation, public outrage, and the very real possibility that PG&E would be broken up or municipalized.
The bankruptcy court had to approve our $130 million retention bonus plan for 14,000 employees—average payouts of $13,000 while 85 families mourned loved ones lost in the fire. Our stock had collapsed 51% in a single day. Employee equity was worthless. Media coverage was relentless. And every senior leader I spoke with carried the weight of knowing their company’s equipment had killed people.
AI can process crisis patterns from historical bankruptcies. It cannot feel the weight of 55 families on your talent team depending on your decisions. It cannot calibrate the difference between necessary urgency and panic-inducing fear. It cannot read the room when bankruptcy attorneys are present for every compensation discussion, or sense when your best recruiter is hours from burnout because they’re fielding calls from employees asking if the company will survive.
Fast forward to March 2020. COVID-19 transformed talent acquisition overnight at US Foods. We didn’t need AI to tell us the restaurant industry was collapsing—we needed experience to pivot our entire recruiting strategy within 72 hours, redeploy our team of recruiters to new priorities, and maintain morale when our requisition load dropped 80% and the future of foodservice distribution was uncertain.
Research consistently shows that the vast majority of transformation initiatives fail, with lack of experienced leadership cited as a primary factor. During COVID-19, I observed firsthand that companies whose executives had navigated prior crises—the 2008 recession, industry disruptions, major restructurings—demonstrated markedly stronger workforce stability than those facing their first existential threat.
The difference wasn’t access to better data. It was leaders who’d already stood in the fire.
I’ve led recruiting transformations that delivered $7M in cost reductions and built talent acquisition functions from scratch during M&A integrations across multiple industries. Here’s what no AI model captures: the political landmines hidden in organizational charts, the informal power structures that determine whether change sticks, and the cultural antibodies that reject even brilliant strategies.
When transforming CHC Helicopter’s talent acquisition, the challenge wasn’t identifying inefficiencies—any competent analysis could do that. The challenge was understanding that aviation safety culture made pilots deeply skeptical of “corporate efficiency initiatives,” that operations managers in remote Norwegian bases wielded more influence than their titles suggested, and that our most critical hire wasn’t the best resume on paper but the candidate who could earn credibility in a 3 AM hangar conversation.
Industry research on transformation consistently shows that initiatives led by executives with direct operational experience in the business unit being transformed significantly outperform those led by external consultants or driven primarily by data analytics. The difference? Experienced leaders know which metrics matter and which are vanity numbers. They understand the human systems that make or break change.
I’ve seen this pattern across multiple transformations: AI excels at pattern recognition across thousands of prior initiatives. But patterns aren’t preparation for the moment when your CFO demands headcount cuts the day before launching your new recruiting technology, or when your best hiring manager threatens to quit because they feel replaced by automation, or when the union representatives push back on efficiency gains that eliminate jobs their members have held for decades.
Peter Drucker famously said, “Culture eats strategy for breakfast.” In M&A integration, culture devours AI for every meal.
Research from Harvard Business Review and other sources indicates that between 50-90% of M&A deals fail to create expected value, with cultural misalignment consistently identified as a leading cause. I’ve navigated multiple integrations where the quantitative due diligence was flawless—headcount alignment, compensation benchmarking, systems compatibility all checked boxes. Then we hit reality.
During one integration, our data showed we should consolidate recruiting teams for efficiency. The numbers were compelling: eliminate redundancy, standardize processes, achieve economies of scale. Experience told me that forcing the acquired company’s entrepreneurial recruiters into our structured corporate processes would trigger an exodus of exactly the talent that made the acquisition valuable in the first place.
We split the difference—maintaining separate teams initially while creating cross-pollination opportunities, then gradually converging approaches as trust built and people saw the value of the combined model. Retention stayed at 94%. A pure data play following the algorithm’s recommendation would have cost us our best talent and destroyed the acquisition’s value proposition.
That decision couldn’t come from a machine learning model. It came from having integrated teams before, from reading the micro-expressions when acquired employees talked about “corporate,” from understanding that entrepreneurial recruiters who’d built their own processes don’t respond well to “here’s how we do it at the parent company,” and from knowing that retention metrics are lagging indicators—by the time the data shows the problem, your best people have already updated their LinkedIn profiles.
M&A research consistently shows that companies with prior integration experience achieve significantly higher success rates than first-time acquirers, even when newcomers have superior analytical capabilities. The reason is straightforward: experience teaches you that acquired employees viewing the parent company as “the enemy,” conflicting values around work-life balance that manifest in retention crises, and unspoken norms around decision-making turn every talent decision into a potential minefield.
AI can flag cultural survey mismatches. It cannot detect that the acquired company’s “flat hierarchy” actually means the founder micromanages everything, or that their “aggressive growth culture” is code for unsustainable burnout rates, or that their top performers will bolt the moment they’re subject to your bureaucratic approval processes.
This isn’t a Luddite argument against AI. In my consulting practice at Chicago Recruiting Forum, I help companies deploy AI-powered recruiting tools that reduce time-to-fill and improve candidate quality scoring. I’ve seen GenAI tools draft job descriptions in seconds that would take recruiters hours. I use predictive analytics to forecast hiring needs and identify flight risks.
But here’s the critical distinction: AI augments judgment; it doesn’t replace it.
The framework I’ve developed over 20 years looks like this:
During the PG&E bankruptcy, AI could have identified which employees had transferable skills attractive to competitors. Experience helped me understand that our most flight-risk talent wasn’t necessarily our highest paid—it was mid-career professionals who feared career stagnation during a multi-year restructuring, and senior leaders who’d weathered PG&E through the 2001 bankruptcy, the 2010 San Bruno explosion, and now this crisis—leaders who simply had nothing left in the tank.
No algorithm would have flagged that second group. They were loyal, highly compensated, deeply experienced. On paper, low flight risk. In reality, emotionally and psychologically exhausted in ways that don’t show up in retention models.
After managing a $16M recruiting budget through crisis, transformation, and integration, certain scenarios cement the primacy of human judgment:
When a talented candidate discloses a DUI from seven years ago during a second-chance conversation, AI sees a risk flag. Experience asks: What were the circumstances? What has their trajectory been since? How does this align with our values around redemption and growth?
I’ve hired candidates with complicated pasts who became our most loyal, high-performing employees—decisions no algorithm would have recommended because the pattern matching says “risk.” But humans understand that people are more than their worst moments, and that someone who’s overcome adversity often brings resilience you can’t find on a clean background check.
When your CEO wants aggressive hiring, your CFO demands cost containment, and your operations team needs quality over speed, AI can model scenarios but cannot navigate the political chess match of aligning leadership. That requires reading power dynamics, building coalitions, knowing when to push back versus when to flex, and understanding that the “right answer” is often the one that enough senior leaders can live with—not the one that optimizes a single variable.
By definition, AI learns from historical data. When PG&E faced wildfire liability litigation, bankruptcy proceedings, criminal investigations, and public safety power shutoffs simultaneously—a genuinely unprecedented scenario—experience became the only compass. Pattern recognition requires patterns. Novel situations require human adaptability, the ability to synthesize insights from adjacent experiences, and the courage to make decisions when there’s no playbook.
When announcing a major restructuring during PG&E’s bankruptcy, employees didn’t want data—they wanted assurance from a leader who’d navigated similar challenges and emerged intact. They needed to see someone who carries the weight of hard decisions, who can acknowledge fear while projecting confidence, who understands that people are terrified and need authentic communication, not polished talking points.
AI can draft the communication. It cannot deliver it with the gravitas that comes from having stood in those shoes before, from having made promises you kept and admitting when you don’t have all the answers, from establishing the credibility that makes people willing to follow you into uncertainty.
At Chicago Recruiting Forum, I help mid-market companies deploy AI to transform recruiting efficiency. But my value proposition isn’t technology implementation—it’s bringing Fortune 150 crisis experience to guide how they use these tools in their specific context.
AI will continue advancing. It will get better at pattern recognition, prediction, and process optimization. But it will never experience the gut-punch of announcing layoffs to people who trusted you, the elation of salvaging a critical hire through a 2 AM phone call when a competitor tried to poach them, or the sober weight of knowing your decision will ripple through families and communities.
The future belongs to leaders who wield AI as a force multiplier for experience-driven judgment:
When I’m helping a mid-market CEO decide whether to invest in AI-powered talent analytics, the conversation isn’t about the technology’s capabilities. It’s about whether they have leaders with enough experience to interpret what the AI is telling them, to know when the pattern it’s matching doesn’t fit their unique context, and to make the human calls that algorithms can’t.
When the next crisis hits—and it will—your organization won’t be saved by the sophistication of your algorithms. It will be saved by leaders who’ve weathered storms before, who can make values-based decisions under ambiguity, who understand that talent strategy isn’t about optimization but about humanity.
I watched Paradise burn from news coverage while managing PG&E’s talent response. I pivoted US Foods’ recruiting when COVID shut down restaurants overnight. I’ve integrated acquired teams who viewed us as the enemy and turned them into advocates. I’ve delivered cost reductions during bankruptcy while maintaining the workforce stability that kept California’s lights on.
Those experiences taught me something no amount of data analysis could: AI is the most powerful tool we’ve ever had for talent acquisition. But tools require skilled hands to wield them.
Experience doesn’t trump AI. Experience directs it, contextualizes it, and overrides it when stakes demand judgment that only comes from having stood in the arena before.
That’s the edge AI will never replicate—and the competitive advantage that will matter most when your next transformation, crisis, or integration tests whether your talent strategy holds or fractures.
