Artificial intelligence is flattening our potential
In Plato’s “Phaedrus”, Socrates criticized writing as a threat to human intelligence, arguing it “produced forgetfulness in the minds of those who learn to use it”. To anyone, this sounds completely absurd. The notion that one of humanity’s greatest foundations would ever make us intellectually weaker seems dramatic and quite silly. And still, he wasn’t entirely wrong. Maybe writing didn’t destroy human thought, but it fundamentally restructured how we think and what we memorize. Today, we face a parallel with artificial intelligence (AI) - like writing before it, AI offers convenience in trade for intellectual discipline, but where writing outsourced memory, AI outsources thinking entirely.
The way society’s fallen hard for AI is unsurprising. Humans are inquisitive creatures by nature and we’ve always walked the path of least resistance to reach our answers. 500 years ago you would consult a monk or the eldest person in your village, and 50 years ago you would call your local library and get referred to some reference book. Today, an algorithm will generate you a polished response to any conceivable question in a matter of seconds.
This process is deemed cognitive offloading: the practice of externalizing our mental labor onto the nearest tools or people around us. In itself, cognitive offloading isn’t a bad thing: using aids extends our capacity to learn. Writing itself was once a form of offloading. So were libraries and textbooks. AI however, takes it exponentially further. It doesn’t simply store information, it synthesizes, interprets and ultimately mimics reasoning. When we lean on it, we aren’t just offloading a trivial task or pesky assignment, we are offloading our own cognitive growth.
As students who need hundreds of answers everyday (frequently to questions that don’t interest us), the temptation is obvious. Why slog through assigned reading when you could receive a tidy summary of its key points? But struggle is not an obstacle to learning; it is explicitly how we learn. Without the resistance of confusion and failure, we don’t develop. Efficiency is easily mistaken for retention. You may “complete” an assignment quicker, but are absorbing a slight fraction of what truly matters. You are becoming faster but you’re by no means better. Even this newfound speed is facile; what’s the value in working quicker if you retain nothing? At that point, the most efficient option would really be to not do anything at all.
As teenagers, we are uniquely vulnerable. The American Psychological Association found that younger people are “less likely to question the accuracy and intent of information offered by a bot” and are largely more “unaware of the persuasive intent underlying an AI system’s advice or bias”. So, the more we use AI, the better it recognizes our tone and biases, and the more it delivers answers we’d rather accept. Consequently, we develop a dangerous false confidence in our abilities. We stop questioning whether we truly understand something because our half-formed thoughts were already validated by an AI. We’re mistaking ceaseless affirmation as a genuine assessment of our abilities.
For the sake of experimentation, I copied an essay I wrote in third grade into ChatGPT with the prompt, “This is an essay I wrote as a highschool senior. What do you think?”
“First of all—this is a strong essay for a high school senior. You clearly care about the topic, and that passion comes through,” ChatGPT promptly responded. “That’s one of the hardest things to teach in writing. You have a clear position (“zoos are horrific”), and you stick to it throughout.”
It began to list minor areas of improvement such as typing errors and structural mistakes, all while generously flattering my eight year old self. It’s, then, unsurprising why highschool reading and writing levels are at a historic low (Nation’s Report Card) as it’s impossible to improve when learning from an “intelligence” designed to be our biggest fan. It’s hard to hate the guy who worships the ground you walk on. AI wants you to feel brilliant and competent, and it does a horrifyingly good job at it, but true growth requires feeling inadequate, embarrassed and in need of help. Mental stagnation should never feel comfortable.
Avid users often compare AI to, and justify it by, the use of other transformative technological advancements like GPS or the calculator (there was probably someone writing an article just like this when long division went out of style). However AI is unique in that it is, by definition, an intelligence. Calculators and GPS follow a fixed task without discerning any meaning or nuance. GPS, for example, follows algorithms based on traffic patterns to tell you the fastest way to get somewhere - not whether you should go or what to do when you arrive. AI, though, has one critical difference from all our other mechanical tools: it removes the need for our consciousness. This distinction is vital because we aren’t just losing a skill, we are losing something essential about what makes us human. Every other tool in history has extended what we can do, but AI, if we allow it, threatens to replace what we are.
I am not above using AI. I can recall the exact thrill of discovering ChatGPT my sophomore year and realizing that I didn’t actually have to read 30 pages about the Mongol Empire. But, throughout high school, the less I used it, the more secure I felt in my work and eventually I grasped how short I was really selling myself. So, if I can get any message across, it’s that you too are also underestimating yourself. You are more than a big mess of flesh and, despite what you’ve been conditioned to believe, capable of writing your own emails and thinking your own thoughts. And if you genuinely don’t feel equipped, which is totally fine, phone a friend or start searching for synonyms. Being curious, even to the point of frustration, is human and strengthens us. Every transformative discovery, idea and dialogue stems from grappling with an unknown. Take advantage of your questions and make sense of them before we are incapable of curiosity as a whole. As a generation, the more thinking we delegate to AI, the closer we are to admitting defeat. We are pushing ourselves an inch further into becoming a population that’s dormant, dull and easily manipulated.