Everybody's Using AI? DuckDuckGo's Founder Has the Data to Say Otherwise

Walk into any conference, investor meeting, or team lunch right now and you’ll hear some version of the same line: “I can’t work without AI anymore.” It’s become the ambient assumption of the entire tech industry. So it’s notable when someone pushes back hard — and even more notable when that someone is Gabriel Weinberg, founder of the search engine DuckDuckGo. His counterargument isn’t a vibe. It’s data.

The trap hiding in “everybody uses it”

Start with a simple question: where does your sense that “everyone uses AI” actually come from? For most people in tech, it’s three places — your X timeline, the newsletters in your inbox, and conversations with people in the same industry. The problem is that all three are the same bubble, packed wall to wall with early adopters.

That’s Weinberg’s core point. The people who use AI the most also talk about it the most. So online discourse makes adoption look dramatically higher than it is. Statisticians have a name for this: selection bias. A loud minority gets mistaken for the whole.

Think of it like a gym. If you only talk to the people who show up every morning, you’ll conclude everyone is working out these days. Step outside the building and the world looks nothing like that.

What the search data actually shows

Weinberg is interesting precisely because he isn’t guessing. DuckDuckGo handles hundreds of millions of queries, and the data he sees suggests the share of users who genuinely lean on AI features is far smaller than the conversation implies.

There’s a crucial distinction buried here: trying something once and using it every day are entirely different behaviors. Plenty of people have opened a chatbot out of curiosity. Far fewer have wired AI into their daily workflow to the point where its absence is an irritation.

This tracks with broader surveys. Across consumer research, the share who say they’ve “tried” generative AI is high. Narrow it to “use it weekly” and the number drops sharply. Narrow it again to “use it as a core work tool” and it shrinks further. The top of the funnel is wide; the bottom is a trickle.

Why the illusion is dangerous

This isn’t a harmless case of people overstating things. The illusion can quietly wreck real decisions.

First, product design goes sideways. Build a service assuming every user is fluent in AI, and most ordinary users will find it confusing and unwelcoming. This is exactly how companies ship AI features front and center, then watch churn climb instead of fall.

Second, investment and strategy wobble. Pour enormous resources in on the belief that the market is already mature, and if real demand is still early-stage, the gap between expectation and reality lands straight on the balance sheet as a loss. It echoes the dot-com era, when everyone was certain shoppers would move online imminently. The direction was right. The timing was years off.

Third, policy and public debate get distorted. Framings like “everyone already uses AI, so regulation is too late” — or its mirror, “the jobs are already gone” — make calm, evidence-based discussion nearly impossible.

Somewhere between hype and reality

Don’t misread him. Weinberg is not saying AI is overrated junk. His message is sharper than that: AI matters, but the speed and depth of adoption are being oversold.

Technology adoption always runs on a lag. The internet did it. Smartphones did it. Cloud did it. Between the moment early adopters fall in love and the moment the mainstream actually folds something into daily life, there’s usually a gap of several years. AI is somewhere inside that gap right now — not at the peak, not on the way out, just in the messy middle.

That’s why a data-backed pushback like Weinberg’s reads less like pessimism and more like a sense of reality. You have to clear away the froth before you can see where the real opportunities are.

So here’s the question worth sitting with. The next time you say “everyone’s using AI,” who exactly is “everyone”? The entire user base — or just the familiar faces around you? Telling the bubble apart from reality starts with that one honest question.