We spend endless hours debating whether AI will eventually outsmart humanity, overtake us, displace the workforce, or trigger an existential crisis. But while we stare at the distant horizon waiting for superintelligence, a quieter, far more immediate event is unfolding directly on our screens. Our own minds could start to atrophy.

Here’s my real worry about AI. It’s not that it’s smarter than us. It’s that it sounds smart, speaks with fluent authority, and we’re too lazy to check if it’s actually saying anything at all. I think many of us experienced it. Ask AI for a genuinely new business idea, a fresh angle on a problem, a real breakthrough. What you receive back almost never contains actual novelty. Instead you get page after page of polished sophisticated jargon. It talks about “ paradigm synchronization,” “synergistic value” and “scalable ecosystem integration.” It sounds brilliant and maybe smart. It feels profound. But pull it apart and there’s nothing there or it’s just an old, generic idea, observation or concept dressed up in fancy words.

This isn’t an accident. It’s how the technology works. A 2024 study out of the University of Glasgow looked at AI text through philosopher Harry Frankfurt’s definition of “bullshit”; a speech that doesn’t care whether it’s true, only whether it sounds convincing. That’s what large language models are often optimized to do. Predict what sounds plausible, not what’s actually true or deep.

Harvard Business School and Boston Consulting Group studied 758 consultants using AI. On tasks that fell outside AI’s actual capabilities, consultants using AI were 19% less likely to get the right answer than those working without it. This pattern comes from over-trusting AI and switching off our own judgment. People mistake confidence for correctness and simply handed their judgment over.

I have seen it time and time again in many areas where the teams used all the corporate buzzwords, now even more using AI, to overcomplicate ideas, projects, or processes. I’m not saying we shouldn’t build our vocabulary and master it. But sophisticated words don’t automatically mean a sophisticated idea sits behind them. Bit by bit, we’re trading critical, first-principles thinking for the comfortable illusion of expertise.

And like that, AI isn’t going to beat us by being smarter. It’s going to beat us by making us stop thinking. And honestly, it’s almost funny; we’re busy worrying about sci-fi robots taking over the world, while we happily copy, paste, and forward AI-generated nonsense to our own teams, one email at a time, forgetting how to think for ourselves along the way.

Maybe none of this is new to you. But unlike AI, I’m not trying to dress this up as a big discovery. Just as a reminder, we still have to do the thinking ourselves.

Sources:

1. The HBS/BCG consultant study (19% figure, “over-trusting AI” pattern) Dell’Acqua, F., McFowland III, E., Mollick, E. R., Lifshitz-Assaf, H., Kellogg, K., Rajendran, S., Krayer, L., Candelon, F., & Lakhani, K. R. (2023). Navigating the Jagged Technological Frontier: Field Experimental Evidence of the Effects of Artificial Intelligence on Knowledge Worker Productivity and Quality. Available at: https://www.hbs.edu/faculty/Pages/item.aspx?num=64700

2. The Glasgow “bullshit” paper Hicks, M. T., Humphries, J., & Slater, J. (2024). ChatGPT is bullshit. Ethics and Information Technology, 26(2), Article 38. https://doi.org/10.1007/s10676-024-09775-5

3. Image: AI-generated