The Dunning-Kruger Effect: How AI is Making it Worse
Have you ever met someone who talks like they’ve got a PhD in everything, but when you dig a little deeper, you realize they barely scratched the surface? That’s the Dunning-Kruger Effect in action—the classic case of people who don’t know what they don’t know.
It’s like reading a social media thread on quantum mechanics from someone who, upon further review, has zero scientific background but confidently explains black holes as if they just wrapped up a dissertation on the subject.
It’s that dangerous mix of ignorance and overconfidence. The less people understand a topic, the more convinced they are that they’ve mastered it. Meanwhile, the actual experts—the ones who’ve spent years in the trenches—tend to be the most cautious. They’ve seen the complexities, the unknowns, and the things they still don’t fully grasp.
Now, here’s the kicker: While I am a big fan of AI, I do believe it’s making the Dunning-Kruger Effect problem a whole lot worse.
AI: The Perfect Fuel for Overconfidence
Artificial Intelligence, in all its glory, has given us instant knowledge—or at least, the illusion of it. Type in a question, and boom, you’ve got an answer. But here’s the problem: a half-baked answer delivered with confidence is worse than no answer at all. I actually shared a post earlier this week here on this very topic, “The AI Advice Trap: Why Context Matters.”
AI-generated content, no matter how advanced, often lacks context, nuance, and real-world experience. It pieces together patterns from existing data, but it doesn’t think, doesn’t understand, and definitely doesn’t care whether you make a terrible decision based on its response.
Yet, because AI sounds authoritative, people believe it. They take half-truths and incomplete data, slap a coat of confidence on it, and suddenly they’re self-proclaimed experts.
See where this is going?
The Recipe for Disaster: AI + Dunning-Kruger
Let’s break this down:
AI gives quick, surface-level answers – People read them and assume they now “get it.”
They skip the deep research – After all, why question something that sounds so certain?
Hey, don’t roll your eyes. This happens all the time. I’m guilty of this myself.
People make decisions based on incomplete knowledge – Sometimes small ones (bad takes on X), sometimes massive ones (misguided business strategies, health choices, or legal advice).
They spread misinformation – And because confidence sells, others start believing them, too.
This is how we end up with people confidently debating complex fields—economics, medicine, law, technology—after skimming an AI-generated summary. It’s intellectual fast food. Easy to consume, temporarily satisfying, but ultimately lacking the nutrients that real expertise provides.
But AI Is So Smart… Isn’t It?
It depends on what you mean by “smart.” AI can analyze vast amounts of data in seconds, generate well-structured content, and even mimic the tone of a seasoned professional. But intelligence? That’s something else entirely.
Think about it like this: A calculator is great at math, but it doesn’t understand numbers. It just follows rules. AI does the same—it predicts patterns and assembles information in ways that look intelligent, but it doesn’t have insight, judgment, or common sense. It doesn’t know when it’s wrong, and worse, it doesn’t care when it’s misleading you.
And here’s the real danger: people assume AI is always right. They trust it blindly, not realizing that it can be confidently wrong—which, ironically, is exactly what the Dunning-Kruger Effect describes in humans.
Real-World Consequences: When AI-Backed Overconfidence Goes Wrong
This isn’t just an abstract problem. We’re already seeing the fallout of AI-fueled overconfidence in the real world:
Misinformation on steroids – AI-generated content is flooding the internet with convincing but inaccurate takes on politics, science, and finance. People believe and share it without question.
DIY medical and legal advice – People are using AI to diagnose themselves or craft legal arguments, often with disastrous consequences.
Businesses making high-stakes decisions based on AI shortcuts – AI tools can be useful, but when leaders make major strategic moves based on AI’s “best guess” rather than expert analysis, things spiral fast.
AI isn’t the problem. The problem is people treating AI-generated content as gospel while skipping the necessary critical thinking.
So, What’s the Fix?
We can’t put the AI genie back in the bottle, but we can change how we interact with it. Here’s how:
Stay skeptical. AI is a tool, not an oracle. Treat it like an assistant, not an expert.
Do the work. If a topic matters, dig deeper. Read books, talk to real professionals, challenge your assumptions.
Embrace uncertainty. The smartest people admit what they don’t know. It’s a sign of wisdom, not weakness.
Fact-check everything. AI can be confidently wrong—don’t let its tone fool you.
At the end of the day, AI isn’t making us smarter. It’s just making it easier to sound smart. The real challenge? Making sure we don’t fall for our own illusions.
So the next time someone confidently rattles off AI-fed insights like they’re an expert, ask yourself: Do they really know what they’re talking about, or are they just another victim of the Dunning-Kruger Effect—this time, powered by AI?
Until next time, keep learning and never stop asking questions.
Mitch Jackson, Esq. | links
This post is free.
But free doesn’t build the future.
Independent journalism only works when people like you choose to lean in—not just with attention, but with support.
If this work matters to you, today’s a great day to take the leap.
$5 a month. $50 a year. For you or gift to a friend.
A small investment in something bigger than all of us.
So many AI experts. We are each entitled to our opinions and perspectives but there are an astonishing number of AI pundits and proselytizers out there. Reminds me of all the armchair virologists during the pandemic.