1. Introduction
In this episode of Impact of AI: Explored, we (James O’Regan and Gerjon Kunst) sit down with Paul Slater to tackle a question we’re hearing more and more: is AI actually making us dumber, even as it makes us more productive?
For decades, IQ scores climbed steadily, a trend known as the Flynn effect — but newer research and real‑world signals suggest those gains may be stalling or even reversing, especially as we offload more of our thinking to digital tools and AI. In this conversation, we explore what that means for how we work, learn, and lead teams today.
2. Meet the Guest
Paul Slater is a strategist and thinker focused on the intersection of human potential and AI, known for asking questions like “Is AI actually making us dumber?” He has spent years exploring how tools, incentives, and information environments shape our ability to think clearly and perform at a high level, especially under pressure.
Through his writing and podcast appearances, Paul connects academic research, behavioural insights, and frontline stories from organizations wrestling with AI adoption, always coming back to a key theme: technology should expand human capability, not quietly take it away.
3. Setting the Stage
AI is no longer a novelty; it’s embedded in how we code, write, research, and communicate. The real question is not “should we use AI?” but “what is our relationship to it doing to our minds over time?”.
In this blogpost, we walk through the most important ideas from our conversation with Paul: what the “reverse Flynn effect” actually is, how cognitive offloading shows up in everyday work, and how to design a relationship with AI that keeps you — not the model — in control of your own thinking.
4. Episode Highlights
- From productivity boost to cognitive offloading
Together with Paul, we unpack how AI tools make it dangerously easy to skip the hard parts of thinking — planning, structuring, doing the first draft — and how that can slowly atrophy the mental muscles we rely on for real problem‑solving. - AI as bicycle vs. self‑driving car
Paul brings in a powerful metaphor: AI as a bicycle for the mind versus AI as a self‑driving car. In one mode, you still pedal and build strength; in the other, you sit back, get comfortable, and gradually lose the ability to navigate on your own.
5. Deep Dive: The Reverse Flynn Effect and Everyday Work
The “reverse Flynn effect” describes evidence that, after a century of rising IQ scores, some populations are now seeing declines in aspects of cognitive performance. Paul connects this to our current environment: constant digital distraction, incentives for shallow output, and heavy reliance on AI to compensate for missing time, focus, or skills.
We link this to patterns we all recognise in knowledge work: AI drafting almost every email, spec, or slide deck; juniors reaching for AI before they’ve tried to reason a problem out; and teams defaulting to “ask the model again” when something breaks, instead of debugging the system themselves. None of this is automatically bad, but multiplied over years it can erode deep understanding, creativity, and the ability to handle genuinely novel challenges.
6. Real‑Life Stories & Examples
In the episode, we explore concrete scenarios that bring this to life:
- Engineering teams who ship faster with AI‑assisted coding, but struggle when production incidents require reasoning beyond what a code assistant can suggest.
- Students and early‑career professionals who can produce polished reports with AI, yet stumble when asked to explain their own reasoning at a whiteboard.
- Leaders who see dashboards full of productivity metrics improving, while noticing that fewer people push back, challenge assumptions, or propose truly original ideas.
Paul connects these anecdotes to ideas like “cognitive debt” and “metacognitive laziness” — the way repeated shortcuts can make us feel efficient in the moment while slowly weakening the underlying skills we think we still have.
7. Key Takeaways
- AI isn’t inherently making us dumber, but uncritical, always‑on use can quietly erode memory, critical thinking, and creativity.
- The real risk is a workflow where you never struggle, never draft, and never think from first principles anymore because AI is always there to fill the gap.
- Using AI as a “bicycle for the mind” means thinking first, then using the model to critique, expand, or stress‑test your own ideas.
- Leaders need to be explicit about where AI can automate and where humans must remain in the loop and in control, especially in high‑stakes decisions.
- On an individual level, we can protect our cognitive fitness with deliberate habits: device‑free thinking time, doing some tasks “the slow way,” and practicing skills without AI assist.
8. Closing Thoughts
Recording this episode with Paul left us with a clear tension: AI is an incredible amplifier for our work, but it can also become a very comfortable shortcut away from the kind of thinking that makes us valuable in the first place. The real choice isn’t “AI or no AI” — it’s whether we design our use of AI to stretch our minds, or let it quietly shrink the space where we actually think.
We’d love to hear how you’re navigating this in your own work. Are you using AI as a thinking partner, or do you catch yourself outsourcing your brain to it? Drop us a comment, share your setup or your rules of thumb, and stay tuned for the next episode of Impact of AI: Explored, where we continue to unpack what it means to stay human in an AI‑saturated world.
