9 of 1,669
AI Can Build the Science, But Not the Scientist
Inbox
| John Nosta via Psychology Today <noreply@psychologytoday.com> Unsubscribe | Wed, Jun 10, 8:40 PM (2 days ago) | ||
to me | |||
| View in web browser A note from John Discovery is key in science. But what will AI do to the person making the discovery? Read it. You might discover something about yourself. John AI Can Build the Science, But Not the Scientist Discovery changes the discoverer in ways algorithms cannot replicate.  By John NostaKEY POINTS:AI accelerates discovery, but the struggle to discover is what forms the scientist. Judgment can’t be transferred, but built through curiosity, error, and revision.The greatest invention may be the mind that makes invention possible.  Source: Image by David from Pixabay. Nikola Tesla once described a curious and remarkable habit. Before building a machine, he would build it in his mind. He would run it, test it, and find flaws without ever touching a physical prototype. The alternating current motor that would eventually help electrify our modern world existed first as an act of Tesla’s mental construction. What interests me isn’t whether this was actually true in every detail. What interests me is the possibility that the process itself was the point. The motor wasn’t simply an invention waiting to be revealed. The time Tesla spent imagining and refining it may have been spent developing something else at the same time. Tesla himself. A recent article in Noema argued that human intuition remains essential to scientific discovery even as AI becomes increasingly capable. I agree. But the argument may stop one step short. The deeper question for me isn’t just if intuition helps produce discoveries. It’s whether the process of discovery helps produce the kind of mind capable of discovering anything at all. We tend to think of knowledge—the theory, the experiment, the breakthrough—as the primary output of science. Yet scientific work produces something else that rarely appears in journals or textbooks. It shapes the minds engaged in the pursuit. A researcher who spends years toiling with a difficult problem develops more than expertise; they develop judgment. A feel for which intuitions deserve trust and which fail under closer scrutiny. They become, in ways that are hard to measure and easy to undervalue, different people than they were when they began.We may be misreading the role of difficulty. Confusion gets treated as the cost, and the answer as the reward. But patience and humility don’t arrive after the struggle; they emerge from it. Being wrong may not simply precede understanding; it might be the mechanism through which understanding and the individual form. Imagine Tesla with access to modern simulation tools. The motor arrives sooner. Design flaws are caught instantly. And years of effort compressed into weeks. From the standpoint of productivity, it’s an obvious improvement. But would Tesla have developed his extraordinary capacity for mental simulation if the tools had done the simulating? Would the mind that produced those answers have formed in the same way if the answers had arrived more easily and earlier? Knowledge and judgment are not the same thing. Knowledge can be transferred, but judgment must be cultivated. One can be delivered while the other must be developed through the very human constellation of curiosity, ambiguity, error, and revision.AI can generate science. That’s the good news and the bad news. What it can’t do is spend years toiling with a problem, getting it wrong, and then becoming someone capable of getting it right. That remains ours. ReferencesGenerative AI lacks the human creativity to achieve scientific discovery from scratch. Scientific Reports. Nature. A. Wenxuan Ding et al. March 2025. John NostaThe Digital SelfTechnology, Transformation and the Future You |
canisgallicus 