Don’t Eat AI Crayons (and Other Tips for Loving AI While Protecting Your Critical Thinking Skills)
In June, researchers at MIT published a study with evidence that certain uses of AI can reduce a person’s ability to think critically. For those of us who love AI, we may want to dismiss these findings as an objection by people who don’t like technology, aren’t interested in innovation, or are rigid traditionalists like my elementary school teacher Mrs. Stewart (who seemed to take pride in making students cry). Conversely, for those of us who are not users or fans of AI, we may be patting ourselves on the back in congratulations for knowing better. My friends, the solution is neither to dismiss the study nor congratulate ourselves for knowing better than to trust AI. Let’s take a closer look.
Don’t Eat AI Crayons
I colored a lot as a kid, particularly with Crayola crayons. One memory stuck in my brain—aside from the smell—is the statement on the front of the box: “non-toxic.” Even as a kid, that reassured me. Crayola cared enough to make sure that one of my favorite art supplies was not going to hurt me. I imagine that if ChatGPT or Gemini came in a box, it would have a non-toxic label too, assuring us that what we would find inside was nothing more than a tool for creativity and productivity. And yes, just like you can use crayons without toxicity, you can do the same with generative AI. But what if I ate a dozen of those crayons? According to The Cleveland Clinic, I would experience a tummy ache and maybe some rainbow diarrhea. Let’s take it a step further. What if I replaced my food consumption with crayons? Well then, yes, despite being non-toxic, I surely cannot nourish myself on a diet of Crayola alone. I would indeed become sick, lose energy, and cease to be able to do the things I once did. Really, it’s the same thing with AI. A little is not going to be toxic. By simply entering a prompt, I am not going to corrupt my brain. But if I start to trade my reading, listening, consideration, thinking, and writing for one-click prompt entry, even the most established neuronal pathways in my brain will begin to atrophy. The problem is, while it’s pretty much common sense to most adults not to eat a box of crayons, AI is new if not entirely unfamiliar to us. It does not come with directions or even generational experience on how to use it while maintaining our own cerebral well-being. Here are a few tips to prevent yourself from binging on metaphoric AI crayons.
Become Fluent With AI
If you can’t explain how your AI tool is working at a high level, take a moment and seek some answers. You could even ask AI to explain it to you! When you understand the basics of AI, you will have a much better sense of which tools to use when, as well as their shortcomings and strengths. Also, before you rely on AI, play with it to see what it can do. For example, when using generative AI like ChatGPT experiment with different prompts around the same topic. Experiment with how different words will yield different results.
Trust But Verify
When you choose to use AI, remember that it is fallible. From transcription to composition, AI makes mistakes. If you trust without verification, you’ll find that AI’s errors will become yours. Check your transcriptions for accuracy, especially when names are used or when there is technical vocabulary. Make sure you vet information from generative AI by knowing the topic ahead of time, checking your AI’s sources, or doing some fact checking to support your AI response.
Reflect
After you’ve done the initial research or reading for your project, take some time to reflect on it by writing why it’s important to you, noting what connections you’ve made between your new information and your previous knowledge, and any asking questions that will lead you to additional information. This step will help your brain begin to move your new information into a place where it can be stored for a longer time. It will also help you analyze your newly acquired information for use in upcoming work, whether your intent is to write, speak, or simply to build upon your general knowledge.
Question
Rather than straight up asking AI for an answer to a question or a problem, offer it a statement and then ask multiple questions about the statement. For example, instead of asking “Is recycling a good program for cities?”, make the prompt a statement and follow it with a set of questions. “Many cities have recycling programs. Are they profitable? Are they effective at reducing landfill waste? How do citizens benefit from recycling? Are the benefits monetary or environmental?” Not only will you examine the statement from multiple angles, but you’ll find that simply thinking of questions will fire up your brain’s active thinking. You can also use this technique to analyze your own thinking by stating your thesis or hypothesis and inviting the AI to poke holes in it. “I think that purchasing a used car is a better investment than buying a new one. Take on the persona of a new car salesman and challenge my assumptions with information I should consider.”
Keep your Brain Toned
Brains are like muscles. When you don’t use them, the load they can carry diminishes. Make sure you do enough of the thinking work on a project to keep your brain in good shape. And if you want to be in great shape, reserve your AI use for when you are truly limited by time. For example, consider doing all of your research on your own. This way, you are learning the information as you go and you will be in a good position to evaluate the accuracy of any AI-generated text you create. Stay in control of your thesis statement and the structure of what you are going to write. This will ensure that your writing reflects what you’ve learned and that your written product will fulfill your intended purpose for creating it. Finally, write the first draft yourself and use AI to polish. This will keep your writing skills in working order, align your AI content more closely with your purpose, and exercise your executive function skills by determining which information is important and in which order it should be presented.
