Hello, fellow humans! On Tuesday, we started talking about where humans fit in the loop, and I outlined a roadmap for developing ourselves to be smarter and wiser with AI, automations, and agents. Today, I have a deeper dive with part 1 of the series and some actionable intelligence and insights.

Every automation hides a decision about values. When we ask an AI to “optimize” a supply chain, we’re not just asking it to run faster or cheaper—we’re choosing which definition of optimal we care about. Is it speed? Cost efficiency? Carbon footprint? Durability? Resilience during disruption? Fairness across vendors? Every one of those metrics encodes a worldview about what is meaningful, what we believe is worth growing, and what trade-offs we’re willing to accept along the way.

Most organizations treat these choices as technical settings—dropdowns in a dashboard. But they’re not technical at all. They’re moral, economic, and strategic decisions about what good looks like in a business. And here’s the part we’re not talking about loudly enough:

AI automates our assumptions.

Before AI agents, many of these value judgments were buried inside human habits, intuition, and tacit knowledge—what we casually called “common sense.” And because they lived mostly in people’s heads, they were inconsistently applied, argued over in meetings, or softened by context and empathy.

But now?

AI gives us the power to codify those assumptions into systems that scale instantly, repeat consistently, and operate relentlessly.

And that’s both the breakthrough and the danger.

The danger isn’t that AI will think for us—it’s that AI will amplify what we haven’t thought through. If our model prioritizes cost above all else, it will happily squeeze quality and resilience until something breaks. If we reward speed, it will race past nuance, ethics, and long-term risk. If we overvalue efficiency, we’ll unintentionally starve exploration and innovation.

This is the hidden cost of automation: not the price of the software, but the price of the implicit worldview we accidentally encode inside it.

Ready to transform artificial intelligence from a buzzword into your personal revenue generator

HubSpot’s groundbreaking guide "200+ AI-Powered Income Ideas" is your gateway to financial innovation in the digital age.

Inside you'll discover:

Download your guide today and unlock a future where artificial intelligence powers your success. Your next income stream is waiting.

Get Your Guide

AI is changing the nature of work faster than organizations can redesign their job descriptions. For decades, the economy rewarded people for their ability to produce: generate reports, analyze data, generate proposals, draft documents, create plans, and manage day-to-day execution.

AI is swallowing those tasks, not all at once but piece by piece. What’s left for human talent is the layer above production:

directing intelligence, not doing the work itself.

And directing requires metacognition—the ability to think about how you think.

This is going to be one of the most important leadership skills of the next decade. Why? Because directing AI systems and AI-assisted workflows forces leaders to ask questions that traditional management never required:

In other words, instead of asking “Can the model do this?” we now have to ask:

“Should it do this? For whom? Under what constraints? And at what cost?”

This shift is especially jarring for professionals whose roles have historically revolved around production—people who have defined their value through output. For them, AI feels less like a tool and more like an existential threat.

But for product managers (and systems thinkers more broadly), this is familiar territory. Product professionals have spent years defining success metrics, evaluating trade-offs, modeling customer value, and coordinating cross-functional decisions. They know that every prioritization framework is really a value framework in disguise.

Now the rest of the workforce has to learn that same muscle.

Let’s be honest: as organizations, we’ve never been great at articulating our assumptions. We confuse inertia with strategy. We operate on tacit rules that nobody wrote down. We rely on mental shortcuts built from anecdotes. We make decisions because “that’s how we’ve always done it.”

In a human-driven workflow, these flaws create friction, misalignment, and inefficiency—but people can recover. A human can notice when something “feels off,” rethink an assumption on the fly, or adjust based on empathy and context.

AI cannot.

AI does not know when a cost reduction harms product quality.

It cannot sense when a customer’s frustration is about underlying trust, not the ticket they submitted.

It cannot intuit that a metric is outdated or narrow or missing something important.

It does exactly what it is told—and it does it at scale.

That’s why metacognition is no longer optional. Our ability to step back, think about how we’re thinking, and choose our metrics intentionally becomes the linchpin skill that separates responsible automation from reckless acceleration.

Here are the practical steps leaders and operators can take to embed metacognition into their teams and workflows:

List out the implicit values behind your KPIs. Ask:

Instead of “Automate this task,” ask:

Shift expectations from “create” to “curate,” from “produce” to “prioritize.”

Every time an AI agent proposes an output, ask:

Verification skills—spotting mismatches, poor trade-offs, or missing context—are becoming the new baseline competency.

Someone must decide which values matter most: speed, cost, quality, risk, fairness, or sustainability.

A one-page “value card” can prevent months of misalignment later.

Prompt it to critique your assumptions, propose alternative metrics, and reveal blind spots.

AI’s real power isn’t in doing work for us—it’s in revealing the assumptions we’ve been making all along.

And the real risk isn’t that it will replace our thinking—it’s that it will scale our unexamined thinking faster than ever before.

The shift from doing to directing isn’t just a technical change.

It’s a cognitive one.

And the leaders who master metacognition will be the ones who thrive in the age of intelligent systems.