Back in the day, a simple fact rang true: understanding required construction. If you wanted clarity, you had to build it from the inside out. You struggled with ambiguity and lived with uncertainty longer than was comfortable. But after all, you were building something. Importantly, this “cognitive building” didn’t appear fully formed but emerged brick by brick and thought by thought. That work wasn’t a barrier to thought; it was the architecture of thought itself. The distance between question and answer shaped judgment because the journey required the brick and mortar of human cognition.
Today, that journey has competition. Artificial intelligence produces structures the same way computer-aided design renders architectural designs. These are not fragments to assemble or clues to interpret. They arrive structurally complete and often persuasive in their accomplishments. And, perhaps more importantly, they integrate seamlessly into our existing mental models. When finality is delivered rather than constructed, something changes in the experience of reasoning. And in today’s world, the word easy shares a border with another equally powerful word, cheap. Let’s examine this.
From Construction to Selection
When we accept that an answer costs less than generating one internally, cognition changes or adapts. The shift isn’t some moral justification but more so as structural. Our minds have commonly conserved effort when lower-cost pathways were available. What’s new is the scale and immediacy of inexpensive coherence. For the first time, this structured completion itself can be externalized. Now, coherence can be selected rather than built. Selected rather than built, that’s the key distinction.
I suggest that this redistribution of cost can alter habits over time. Fluency can begin to substitute for depth because it arrives without struggle. Confidence arrives before construction itself because the language feels complete and authoritative, and the gap between question and closure narrows. None of this reflects diminished intelligence; it reflects an environment in which the “energetic balance” between construction and selection has shifted.
Why Effort Now Feels Different
Maybe the following can explain why discipline alone often feels insufficient. In the context of AI, we aren’t resisting temptation on neutral terrain. We’re navigating a cognitive landscape where the path of least resistance has moved. Human thought remains possible, but it now requires consciously choosing the higher-effort route. The decision is rarely dramatic because it unfolds incrementally. Today’s challenge is whether we accept AI’s fluent answer or sit with the cognitive tension a little longer.
Adaptation, Not Decline
Every major cognitive technology has redistributed mental effort. Writing externalized memory, calculators externalized arithmetic, and search engines externalized retrieval. AI externalizes structured completion itself, the final step before acceptance. When technology changes what costs effort, minds reorganize accordingly.
The question isn’t whether to use artificial intelligence, but whether we recognize the shift it introduces. When coherence becomes inexpensive, construction itself becomes optional. Capacities that are not exercised may gradually recede from habit. Our ability for deep and introspective thinking remains intact, but I contend that it no longer sustains itself automatically. In a world where answers arrive pre-built, depth requires awareness and, at times, the willingness to choose the more demanding path.
These ideas are developed more fully in my new book, The Borrowed Mind: Reclaiming Human Thought in the Age of AI.