Why Every AI Is Telling You to Read the Same Book
When you ask AI for a book that will change your life, there's a 73-in-100 chance it tells you Man's Search for Meaning. When you ask it in a new chat, same odds. When your competitor asks in a new chat, same book. Same advice. Same everything.
A recent experiment tested this across 65 different AI models on simple questions—a color, a city, a career path. Averaged across everyday decisions, 10 to 50 asks each, the chance that two separate new chats with the same AI give identical answers? Between 10 to 50 percent per question. For advice on major life decisions, it's worse: on half of ten tested decisions, the AI gave everyone the exact same recommendation.
This isn't a bug. It's how language models work. They predict the next most likely word based on everything they've seen on the internet. When millions of people ask the same question, they're asking the model to output the statistical average of what's already been written. The model obliging isn't it being smart. It's it being average.
Why This Breaks Strategy (and Business)
Harvard researchers tested six major AI models on business strategy. Every single one recommended the same playbook: differentiate, collaborate, think long-term, augment your capabilities. Word for word, model after model. Same advice. Different logos.
They called it trendslop—safe, generic, middle-of-the-road advice that could apply to any company in any industry. The problem isn't that it's wrong. It's that everyone got it. When you use AI for your strategy, your product, your story, and expect a different outcome—you got the same outcome your competitors got. This morning. With the same AI model.
For Creators, It's a Warning
One study showed that when consultants used AI, variation in their recommendations dropped significantly. Everyone started sounding the same. The AI-assisted stories were judged as more creative at first, but only because the least-creative writers got a boost to baseline. The most creative were held back.
Frontier models like Claude and GPT consistently hit the same answers because they learned from nearly the same training data: everything on the internet. Let it do the driving, and you're buying exactly what everyone else bought.
The Workaround: Be Specific, Be Weird
Some researchers found what breaks the pattern: give the AI something nobody else gives it. Specific details about you. A number to choose with. Constraints that force it off the common path. Ask for the uncommon answer, and everyone still gets the same uncommon answer—but at least it's not the first one.
For creators, the consensus among communities is blunt: the most profitable creators in 2026 will be unmistakably human. Be messy. Be you. The consumer is getting savvy—they can feel the robot. Your weirdness is your wealth.
AI isn't a tool for standing out. It's a tool for catching up. If you're using it the way everyone else is, you're converging. The move is to use it as a starting point, then violently disagree with it. Fight the default. Add what only you can add.
What's Actually Good
AI advice is consistent, which some people want. It's cheaper than a consultant. It doesn't waste your time with bullshit. The problem is it wastes your time with everyone else's bullshit.
The real limitation: AI can't see the edges where your specific problem lives. It can't tell you what your competitor doesn't know because it only knows what's been written. For creators, that's not a limitation. That's your opening.