Why AI can't replace good judgment
Reid Hoffman called AI a "consensus machine." Here's why that's great news for smart lawyers.
Recently, LinkedIn founder Reid Hoffman called AI a “consensus machine.”
He’s right. And that’s exactly why good lawyers won’t just survive—they’ll dominate.
The consensus trap
AI excels at finding patterns. It spots what most people believe. It summarizes the conventional wisdom.
But legal thinking? That’s not about following the herd.
It’s about knowing when to break from it.
A lesson from the bench
I learned this clerking for Judge Duplantier in the Eastern District of Louisiana.
I once wrote a memo urging him to grant a motion in limine to exclude evidence. Solid research. Plenty of supporting case law. Airtight legal analysis.
He skimmed it and said, “I’m not gonna do that.”
I stammered: “But precedent says—”
He cut me off. “How much trial time does this save?”
I hadn’t considered that.
Real analysis
Judge Duplantier had, and he explained his policy for motions in limine:
Grant them only if they’re (1) legally appropriate, AND (2) save significant trial time.
His reasoning? Excluding evidence increases reversal risk on appeal, leading to a retrial. Why create that risk unless it meaningfully streamlines the trial?
No case law discussed this approach, and the opposing lawyer didn’t raise it either.
It wasn’t legal analysis. It was judgment based on experience.
AI shortcomings
AI would’ve made the same recommendation I did, citing the same cases—probably faster and more comprehensively.
But it would’ve been equally useless to Judge Duplantier.
Why? Because AI operates on consensus. It tells you what most courts do and what most lawyers contend.
But great lawyers don’t think like most people.
The important questions
Great lawyers ask: What’s the trade-off? What’s the real impact? What matters most in this situation?
These aren’t questions AI can answer. They require judgment from experience, not pattern recognition.
The uncomfortable truth
If your work relies on superficial reasoning—finding cases, summarizing rules, following precedent—AI will replace you.
And frankly, it should.
But if you can see beyond the consensus? If you can think strategically about outcomes, risks, and real-world consequences?
You’re not just safe. You’re invaluable.
The dividing line
Most lawyers are consensus machines. They do what most lawyers do, argue what most lawyers argue, think like most lawyers think.
AI will eat their lunch.
Lawyers who think differently—who ask better questions, see beyond the obvious, and focus on what matters—will thrive.
The choice
In a world where AI can instantly generate legal research and draft standard documents, your value isn’t in knowing what everyone else knows.
It’s knowing what they don’t.
The question isn’t whether AI will change legal practice; it’s already happening.
The question is: What type of lawyer are you?
;-)
Ernie
P.S. Want to develop better judgment alongside other good lawyers? Join my ChatGPT Lab (a weekly workshop in the advanced level of my Inner Circle). That’s where we discuss new challenges and opportunities in an AI world.
Try it for 30 days for $1 (if you continue, you’ll get a big discount off the regular subscription).