For most of history, intelligence was a scarce resource. If you could think faster, remember more, and solve complex problems better than others, you had an edge. Schools and careers were built around this advantage. But that world is gone.
Now, with advanced AI, especially large language models (LLMs), has changed the equation. Intelligence, at least in the traditional sense, is no longer the bottleneck.
We now have AI systems that can process information faster than any human, write better than most professionals, and pass graduate-level exams in multiple disciplines.
The result? The value of raw intelligence has plummeted. The new bottleneck—the thing that differentiates those who succeed from those who get left behind, is no longer intelligence. It’s judgment.
The Decline of the Intelligence Premium
For decades, being smart was the surest way to get ahead. Smart people were good at retaining information, reasoning through problems, and making connections faster than others. Our entire educational system was built around this assumption. Students were rewarded for knowing facts, following logical steps, and finding the right answers.
But LLMs have shattered this model. AI can now answer almost any factual question instantly. It can summarize books in seconds, debug complex code, and even generate new ideas on demand. The sheer computational power of AI means that intelligence, in the traditional sense, has been automated.
If AI can outperform humans at intelligence-driven tasks, what’s left for us?
Judgment - New Differentiator
AI is great at answering questions, getting better at reasoning and so on, but terrible at knowing which questions to ask.
It can optimize, but it can’t decide what to optimize for. It can generate content, but it doesn’t know what’s worth writing about.
This is where humans still have an edge. Judgment, the ability to prioritize, to discern signal from noise, to ask the right questions, is still uniquely human.
In a world where AI can give you a hundred plausible answers, the real skill is knowing which one actually matters.
Think about investing. AI can analyze financial reports, track market trends, and run sophisticated simulations. But it can’t predict human psychology. It doesn’t know when a trend is actually meaningful or just noise. The best investors don’t just process data well—they see things others don’t.
Or consider startups. AI can generate business ideas, write marketing copy, and optimize user engagement. But it doesn’t have taste. It doesn’t know whether a product is actually compelling or just statistically likely to succeed. The best founders have a feel for what people actually want, often before those people know it themselves.
The people who will thrive in this new world aren’t necessarily the smartest. They’re the ones with the best judgment.
How Education Needs to Change
If judgment is the new bottleneck, education is completely misaligned with reality. Schools are still optimized for an era where intelligence was scarce. They train students to memorize information, follow instructions, and solve well-defined problems.
But what’s the point of memorization when an AI can retrieve any fact instantly? Why train people to solve problems that AI can solve better?
Education needs to shift from answering questions to asking them. From optimizing for correctness to optimizing for insight.
Here’s what that might look like:
1. Train People to Ask Better Questions
Right now, education rewards students for answering questions correctly. But in a world where AI can generate perfect answers, the skill that matters is knowing what to ask in the first place.
Imagine a history class where students aren’t graded on their ability to recall dates but on the depth of the questions they ask about historical events. Or a science class where students are rewarded for designing interesting experiments rather than solving pre-defined equations.
2. Shift from Knowledge Retention to Judgment Development
Knowledge used to be a competitive advantage. It no longer is. What matters now is the ability to think critically about knowledge.
Future education should focus on developing taste—knowing what information is worth paying attention to. This means exposing students to ambiguity and forcing them to make decisions with incomplete information. AI can provide the data, but humans need to learn how to navigate uncertainty.
3. Move from Optimization to Exploration
Most education today is built around optimizing for correctness—getting the highest score, the best grade, the most efficient answer. But judgment doesn’t work that way.
Judgment comes from exploration, trial and error, and intuition. Instead of giving students a rigid curriculum with fixed answers, we should encourage them to explore. What happens if they take an unconventional approach? What if they pursue ideas that seem wrong at first but might turn out insightful?
AI can optimize. Humans should explore.
The New Career Advantage - Humans Who Use AI Well
If intelligence is no longer scarce, what kind of people will be most valuable?
The answer is people who use AI as a force multiplier for their judgment.
The best investors, founders, scientists, and writers of the future won’t necessarily be the smartest in a traditional sense. They’ll be the ones who:
- Ask the best questions (because AI can answer them).
- Filter out noise (because AI generates an overwhelming amount of information).
- Recognize original ideas (because AI is good at remixing but bad at knowing what’s novel).
- Think in ways AI can’t (because AI is trained on existing patterns, while humans can invent new ones).
Success won’t go to those who compete with AI—it will go to those who orchestrate AI.
The Risk of Poor Judgment in an AI-Driven World
Judgment isn’t just about making good decisions—it’s also about avoiding bad ones. In a world where AI can generate thousands of plausible ideas, the real failure mode is picking the wrong ones.
We’re already seeing this. AI-generated misinformation spreads because people don’t know how to evaluate sources. Algorithm-driven decisions fail because companies optimize for the wrong metrics. People fall into intellectual dead ends because they trust AI’s answers without questioning their assumptions.
The danger isn’t that AI will become too powerful. The danger is that humans will become overly dependent on it—losing the ability to think critically and make independent judgments.
If we don’t prioritize teaching judgment, we’ll end up in a world where people blindly follow AI, unable to differentiate between what’s meaningful and what’s just noise.
The Future Belongs to Those Who Think for Themselves
For most of history, intelligence was the bottleneck, so we built education systems to optimize for it. But now that intelligence is abundant, judgment is the scarce resource.
The winners of the future won’t be those who can process information the fastest—that’s AI’s job. The winners will be those who know what’s worth processing in the first place.
In a world where AI can do the thinking, the real skill is knowing what to think about.