When I came into this professional world in the mid-1990s, no one ever sat me down and said, “Here’s how you develop judgment.” That wasn’t a thing. Judgment wasn’t taught. It was acquired — usually under pressure, usually after something had already gone sideways.
We didn’t call it judgment back then. We called it experience.
But if I’m honest, most of what shaped me over the last three decades had very little to do with the visible outputs of the job. It wasn’t the press releases, the media lists, the campaign plans, or the rewrites that mattered. Those were table stakes.
What mattered were the operating conditions those activities put me into: watching narratives form without my permission, realizing too late that silence had operational consequences, seeing how credibility actually moves through an organization or a political system, or owning a decision that affected customers, employees, regulators, or investors and couldn’t be undone once it was in motion.
That’s how judgment used to get built — almost accidentally — and that apprenticeship largely no longer exists.
AI now does much of the work that once created those learning moments. Drafts are instant. Analysis is cheap. Scenarios multiply. Speed is assumed. My instinct is still to preserve the old tasks because they feel like they “build muscle,” but if I do that uncritically, I miss the point—and I train people for a version of business and communications that no longer exists.
If judgment used to be learned accidentally, it now has to be developed deliberately.
How judgment was really built
Looking back, what actually made people effective early in their careers wasn’t mastery of deliverables. It was exposure to pressure inside real operating systems. You learned — often quickly — that narratives don’t wait for approval, that silence has downstream effects, that some voices carry decision-making authority and others only advisory influence, and that credibility is often borrowed long before it’s earned. You learned that the response itself can reshape the business problem, that timing affects trust, cost, and risk — not just perception — and that accountability mattered because there was no undo button.
None of that came from training programs. It came from being close enough to consequence to feel it.
Over time, what we really developed were a small number of durable instincts: the ability to anticipate what will happen if nothing is done; the ability to recognize who actually has influence and authority in a given moment; the ability to think past the first move and anticipate operational, reputational, and stakeholder reactions to the response itself; the ability to know when speed improves outcomes and when restraint preserves trust; and the ability to own results rather than just produce recommendations.
Those instincts haven’t been made obsolete by AI. If anything, they matter more now.
What AI can’t replicate
AI is exceptionally good at mechanics. It drafts, summarizes, models scenarios, and analyzes faster than any team I’ve ever led. What it does not do is decide. It does not prioritize enterprise risk. It does not understand power, fear, incentives, or organizational context the way humans do. And it does not live with consequences.
The risk isn’t that early-career professionals will rely on AI. You should. The real risk is that you begin to confuse speed with maturity, fluency with judgment, and output with leadership. If organizations don’t intentionally create ways for people to experience pressure, tradeoffs, and accountability, they will end up with leaders who sound polished, move quickly, and haven’t developed the instincts that actually protect the business when it matters.
Hiring for judgment, not just output
This is why hiring — and especially hiring recent college graduates — has become so consequential. After all these years of working at the intersection of communications, marketing, government relations, legal, and business operations, I’ve learned that when we are hiring new graduates, we’re not really hiring for output. We’re hiring for trajectory. And in the AI era, that distinction matters more than ever.
AI makes it remarkably easy to sound finished. Answers are structured. Language is confident. Thinking appears clean. None of that tells me much about how someone will perform when priorities collide, information is incomplete, and the business is under pressure.
So the job of an interview isn’t to reward fluency. It’s to see past it. I listen to how candidates respond to uncertainty, challenge, and shared outcomes. When your first instinct is questioned, do you become defensive — or more thoughtful? Do you take ownership of results that didn’t go as planned, or do you explain them away? Blame stops growth faster than failure ever will.
The idea of ‘judgment velocity’
At this stage of your career, I’m not looking for confidence. I’m looking for what I think of as judgment velocity: how quickly you learn as the complexity and stakes increase.
Early careers aren’t shaped by a few visible wins nearly as much as they’re shaped by patterns. People remember who made things clearer, who made decisions easier, who reduced risk, and who made fewer things worse. Reliability compounds. Recklessness does too.
In part two on May 27: how to teach yourself judgment on purpose — without waiting for authority, or a crisis, to do it for you.
David J. Chamberlin is the managing director of the Strategic Communications Advisory Team at Orrick, where, alongside Orrick’s lawyers, he advises clients on reputation risk, communications strategies to address those risks, and global business operations issues. He previously served as the head of global communications at Nortel Networks, the chief communications officer at PNC Bank, and the chief marketing officer at SonicWall.
Illustration: DesignHunt
