The Girl Who Solved the Case
When I was growing up, Inspector Gadget was one of my favourite cartoons. I loved the gadgets that popped out of his hat and wondered what it would be like to have them in real life. But my favourite was Penny and her dog Brain. Penny was one of the few female characters in the shows I watched, and the only one I remember being portrayed as smart and capable. She solved Gadget’s cases without any fancy technology… just her smarts!
If you remember the show, you remember the watch she used to communicate with her dog, Brain. Flash-forward to when I first answered a call on my smartwatch; I definitely felt like a real-world Penny!
If you watch the show now, it is easy to see that Penny never gets the credit. Gadget takes the credit every time and is too oblivious to notice Penny has helped. As a child, I did not notice; it all seemed part of the schtick. It was normalized. This is an unnerving pattern, and it is becoming too common with the use of AI.
Confident, but Wrong
In the article “ChatGPT Is Just an Automated Mansplaining Machine,” Harrison Dupré (2023) argues that ChatGPT is often wrong but always sure it is right. When the article’s writers pushed back with the correct answer, the bot responded with condescending, sassy comebacks. The point that stayed with me most in the article was the quote by AI researcher Gary Marcus, who warned that because AI mimics huge amounts of human conversation, it can “easily fool the uninitiated.” That line struck a chord with me because I am realizing that AI learns from everything our society writes, including all of our unconscious biases. In the same way that society normalized a man taking credit for a girl’s work while I watched Penny outthink everyone, AI is now learning and repeating those same patterns, and we may not think twice about it.
Where Are Today’s Pennys?
If Gadget is like the confident AI, who are today’s Pennys? Hardly any, it seems. I see children and adults taking AI-generated work and answers at face value, with very little critical evaluation. This is the real danger: not just that AI can be wrong, but that very few people are questioning it. Even more dangerous is that students are not being taught the critical thinking skills they need to see it.
Students cannot learn to question what they never see questioned, and as educators, myself included, we can unknowingly pass this problem along. When a tool can ease our mental workload by producing a polished answer, image, or worksheet in seconds, it becomes very tempting to share it without stopping to ask: who is missing from this resource? Does it reflect the students sitting in front of me? This is not done with any intention of harm. Rather, the tool’s confidence makes it easy to stop asking questions, and our students never see us ask them. If we want our students to become today’s Pennys, they first need to see us act like one.
An Old Problem in a New Gadget
The article makes a fair point, but I don’t think it goes far enough. Harrison Dupré treats ChatGPT’s confidence like a new problem, but it isn’t new at all. Wachter-Boettcher (2018) shows that technology has been treating one group of people as “normal” for decades, as when Kodak film set the “normal” skin tone for printing photos using pictures of white women… and no one questioned it. Wachter-Boettcher also explains that sleek, simple designs make biased tools appear neutral, much like Gadget’s confidence fooled everyone around him. What AI adds is a voice that sounds even more sure of itself, and that voice does not reach every student equally. Robinson et al. (2015) found that digital skills are unequally distributed and that these gaps can exacerbate existing inequalities. This suggests that students with fewer digital skills are less equipped to question what a tool tells them and that AI’s biases are likely to hit them harder. Yet, removing AI from our classrooms would not make the bias disappear; rather, it would leave students with the least experience to face it alone. AI can still support many learners, so the answer is not banning it. The answer is teaching kids to question it.
Raising a Generation of Pennys
So where does this leave us? What needs to happen is that we raise a generation of Pennys: kids who think for themselves and use AI as a tool, rather than trusting the gadget to do it for them. I would love for my students to:
- Ask, “How do I know this is true?” and check another source
- Compare the AI’s answer with their own thinking first
- Notice who or what is missing from an answer or image
- Spot when an answer sounds sure but doesn’t explain why
Addressing these inequalities also means changing our own practice as educators: checking what a tool produces before sharing it, and planning with the students most affected by these gaps in mind, offering support or alternatives to those with less access to technology or less experience with it.
I think we must agree that, as teachers and as a society that cares about our next generations, we can’t just hand students these tools and hope for the best. Rather, we must commit to modelling how to question AI by building it into daily classroom habits and into lessons.
Like Inspector Gadget’s gadgets, AI will sometimes malfunction when we depend on it most; the hard work of solving the problem still belongs to the thinker, so our students need to learn to question, test, and verify what a tool gives them, rather than handing the work over to a tool that may fail them without their knowing.
References
Harrison Dupré, M. (2023, February 8). ChatGPT is just an automated mansplaining machine. Futurism. https://futurism.com/artificial-intelligence-automated-mansplaining-machine
Robinson, L., Cotten, S. R., Ono, H., Quan-Haase, A., Mesch, G., Chen, W., Schulz, J., Hale, T. M., & Stern, M. J. (2015). Digital inequalities and why they matter. Information, Communication & Society, 18(5), 569–582. https://doi.org/10.1080/1369118X.2015.1012532
Wachter-Boettcher, S. (2018). Technically wrong: Sexist apps, biased algorithms, and other threats of toxic tech. W. W. Norton & Company.