Intelligence Was Already at the Kitchen Table

Thinking First in the Age of AI and AR

By Billie Powers

Before AI was sitting in a search bar, intelligence was already sitting at the kitchen table. Before a student could type a question into a machine and receive a polished answer in seconds, there was Mom leaning over the homework, Dad explaining the math problem, an older sibling who had already taken the class, or a neighbor who was “good at English.” In many homes, help with schoolwork was part of family life. A parent slowed the child down, asked questions, explained the problem another way, and helped the child find the answer instead of simply handing it over. Sometimes impatience and frustration led to someone else just “doing it for them.” That sounds a little like AI, does it not?

As a mother, I stayed up many nights helping finish school projects, and any parent who has faced poster boards, book reports, missing supplies, and a tired child at the table knows what that means. Sometimes the child needed direction; sometimes they needed someone beside them long enough to get through the frustration. I would have welcomed today’s AI tools, not to do the work, but to help us get unstuck and finish with less exhaustion.

There has always been a point where help can cross the line. A parent may begin by correcting one sentence and then rewrite the whole paragraph. A child may start with a question and end with a project that sounds far more grown up than anything the child understands. The assignment is turned in, the grade comes back, and everyone moves on as though the work came from the student’s own mind. AI did not create that problem, it just made it faster, cleaner, and easier to hide.

The issue, then, is not simply whether someone used AI. The better question is whether the person did the thinking first. Did they read the material, research the subject, sit with the hard parts, and form their own thoughts before asking a tool to help shape the words? That is the line that matters. There is a difference between using help to strengthen your work and using help to avoid the work altogether.

That is where a thesis matters. A real thesis is not merely a sentence placed in the first paragraph because an assignment requires it. It should be the result of thinking. It may begin as an opinion or question, but it must be tested through reading, comparison, doubt, confirmation, and sometimes a change of mind. By the time it is written, it should say, “This is what I believe after doing the work.” That is very different from asking AI to create the position before the person has formed one. 

When the tool writes the thesis first, the student may spend the rest of the paper decorating an idea that was never really theirs. When the person does the work first, AI can help shape the language around a conclusion they already earned. The result should not be a polished page hiding an empty process. It should be a message formed from learning, judgment, and responsibility. The same goes for writing legal papers, books and stories on social media. Just do the research!

Research is already showing why this distinction matters. UNESCO’s AI competency work emphasizes responsible use, human agency, and critical thinking rather than blind dependence. A study published in the Proceedings of the National Academy of Sciences found that generative AI without proper guardrails can improve short term performance while harming learning when students later have to work on their own. That confirms what many parents and teachers already know: help that removes the struggle may look successful in the moment but leave the learner weaker when the help disappears (UNESCO; Bastani et al.).

Someone recently offered to rewrite a legal draft I had prepared because it needed proper legal form and wording. On the surface, that sounded helpful. When I read what came back, however, it was inaccurate, and the intent had been lost. I called to ask questions and explain what the draft was meant to say. While I was still explaining the purpose and substance behind it, they said excitedly, “Look in your email.” That was the moment I lost trust. They had not read deeply, listened carefully, or understood what they were handling. They had simply let AI loose on something that required human responsibility. When speed replaces understanding, trust is often the first thing lost, and in legal work, the consequences can reach a courtroom.

Augmented reality, or AR, adds another layer to this problem. AI challenges us to question what we read and produce; AR will challenge us to question what we see. As these tools become more common, children and adults must learn that a realistic image, convincing video, digital overlay, or immersive experience is not automatically true. AR has great promise when used well. It can help students see anatomy, history, science, engineering, art, or geography in ways a flat page may not. Studies show that it can support learning gains, but engagement is not the same as understanding. A student can be impressed and still not know what they learned. Something vivid is not necessarily true, and something interactive does not automatically create wisdom (Garzón and Acevedo; Bhat, Verma, and Craig).

The same warning applies to AI. Creating a picture with AI or using it to shape a story, organize a thought, or clean up a draft is not the problem. The problem begins when a person claims the finished product as though no tool, teacher, source, helper, or education contributed to it. If someone claims to be an investigative journalist, they should do the investigation. If someone claims to be a book writer, the foundation of that book should come from their research, notes, interviews, lived experience, and formed opinions. AI may help with structure or language, but it should not replace the work that gives writing its truth.

At the same time, people should not be shamed for using tools that help them reach their own thoughts. Some students have learning differences, disabilities, trauma, language barriers, or anxiety around writing. For them, AI can be a bridge that helps organize ideas or explain instructions. But a bridge should still lead a person to their own understanding. It should not replace the crossing.

Teachers are now being asked to judge not only the final paper but also the process behind it. That is difficult when AI detectors are imperfect and human writing can be falsely labeled artificial. Schools should be careful about relying on detection alone. A polished paper does not always mean cheating, and a messy paper does not prove deep thought. A better answer is to examine the trail of thinking: notes, rough drafts, source choices, revisions, questions, and the student’s ability to explain the argument. A person who did the thinking can usually talk about it (Hyatt et al.).

This is why live tutors, teachers, and human advisors still matter. A good tutor can see frustration before the student says a word, hear hesitation, notice when confidence drops, and adjust with empathy. My son-in-law teaches and tutors Chemistry. I respect his methodical, observant, purposeful mind. We just watched a series together, and I enjoyed asking him what I had missed or not fully understood. We shared perspectives, connected the dots in the story, and I came away seeing things I had overlooked. It reminded me why learning with another person is so valuable. A machine can explain a problem, but it cannot care whether the person on the other side feels defeated, overwhelmed, or ready to try again. 

This problem is not limited to students. Adults love shortcuts too. Professionals use AI to sound more prepared than they are. Businesses use it to write messages that sound sincere when no one has taken the time to be sincere. Social media users share polished posts, dramatic videos, and convincing images without asking where they came from or whether they are true. Stanford’s work on misinformation shows that students can be taught to spot questionable sources, while Pew Research Center has reported concern about AI’s effect on creative thinking and human connection. Research on AI generated misinformation on X has also found that synthetic misinformation can spread in ways that should make all of us more cautious before sharing what looks convincing (Stanford Graduate School of Education; Pew Research Center; Drolsbach and Pröllochs).

A society that stops thinking for itself becomes easy to lead, fool, and control. AI can sound confident when it is wrong. AR can make something look real when it is not. These tools should force us to become more disciplined, not less. We cannot let polished words replace truth or realistic images replace judgment. Responsibility still belongs to the human being using the tool, reading the answer, sharing the post, publishing the article, teaching the child, or claiming the work.

Mom and Dad’s, siblings and others were the original intelligence at the kitchen table, and some nights that intelligence was running on fumes, coffee, and pure panic, but the lesson has not changed. AI did not invent shortcuts, and AR will not invent deception. In an age of instant answers and manufactured reality, the question remains: Did you use your own intelligence to learn, or did you borrow it from AI, AR, or another person and call it yours?

Be the intelligence at the kitchen table. Let AI and AR help refine your work, but never let them replace the conversations, questions, and human judgment that build real understanding.

Resources

Bastani, Hamsa, et al. “Generative AI without Guardrails Can Harm Learning.” Proceedings of the National Academy of Sciences, 2025.

Bhat, K. Raghav, Vipin Verma, and Scotty D. Craig. “Mobile Augmented Reality Impacts Engagement, but Not Learning.” Computers & Education: X Reality, 2025.

Drolsbach, Chiara, and Nicolas Pröllochs. “Characterizing AI-Generated Misinformation on Social Media.” arXiv, 2025.

Garzón, Juan, and Juan Acevedo. “Meta-Analysis of the Impact of Augmented Reality on Students’ Learning Gains.” Educational Research Review, vol. 27, 2019, pp. 244–260.

Hyatt, J. P. K., et al. “Using Aggregated AI Detector Outcomes to Eliminate False Positives.” Advances in Physiology Education, 2025.

Pew Research Center. “How Americans View AI and Its Impact on People and Society.” 17 Sept. 2025.

Stanford Graduate School of Education. “It Doesn’t Take Long to Learn How to Spot Misinformation Online, Stanford Study Finds.” 19 Apr. 2022.

UNESCO. “What You Need to Know about UNESCO’s New AI Competency Frameworks for Students and Teachers.” 20 Feb. 2025.

www.tollandroll.com