AI Is A Digital Super Soldier Serum
So I saw this and it gave me thoughts.
I have found precisely the opposite. Using ChatGPT as a thinking partner has sharpened and deepened my thinking, making me demonstrably smarter. So I think this hand-wringing, which is just the latest in a long line of doomsaying every time a new technology comes along (“With these digital calculators, kids won’t even know how to use slide rules anymore! Then what?”), misses the point.
The key distinction here isn’t the technology, but how it’s used. In the case of AI, “do this with me” versus “do this for me.” Engaging actively with AI helps you interrogate your own thoughts and strengthen your argument, where—as the research implies—offloading your work entirely to the AI gives you less to do and encourages mental atrophy. It says more about the human than the technology. Are they the sort of person that wants things done for them, or who approaches life with intellectual curiosity and a drive to improve? In short, AI makes smart people smarter and dumb people dumber.
Think of AI as a digital version of Erskine’s Super Soldier serum. It just amplifies what’s already there. Steve Rogers was already brave, selfless, and determined before the serum; it just gave him the physical strength to match his character. Meanwhile, Johann Schmidt (Red Skull) was already cruel and power-hungry, and the serum just amplified those traits to monstrous levels.
AI operates similarly:
If you’re intellectually curious, AI becomes a powerful tool for exploration, insight, and skill refinement.
If you’re disengaged or reliant on shortcuts, AI makes it easier to avoid thinking critically.
It’s not the tool that’s good or bad—it’s how it’s used and who is using it. And just like with the serum, the real key is ensuring it’s in the right hands (or at least encouraging people to use it in ways that promote growth rather than stagnation).
Oh I totally agree with this and “AI makes smart people smarter and dumb people dumber” nails it. Eight months into experimenting with both ChatGPT and Claude, I’ve learned I actively hate when they get too involved or try to take over. I particularly resist involvement with my writing except to point out my actual errors, like an elaborate spellchecker and impulse/dyslexia monitor, coming to AI last not first in the whole writing process.
What I use them for is to stretch my own thinking much further (and much quicker…I feel like I’ve made over a decade’s headway in my self-understanding as a person with complex neurology and health in those first few months) and to minutely log my ongoing health experiments (otherwise known as life) in real time so that patterns start to emerge with symptoms, causality and treatments/adjustments.
Face it, nobody else is going to chew the cud with me on these highly subjective topics, meaning I’ve spent decades lacking the kind of pushback that grows intelligence and wisdom. At last, instead of losing myself down long winding avenues of abstraction, getting nowhere quickly, I can feel something else generating resistance or flow at the other end of every random thought I throw out there, vastly broadening the map.
Exactly. The introspection I've done since diagnosis wouldn't have happened without ChatGPT as a sounding board.