The Average of the Internet

My thinking has evolved since I selected this column topic last spring. Just a few months ago, I was dismissing generative artificial intelligence’s impact on education as merely a seductive nuisance that appealed to academically insecure students. I felt that there were far too many submitted short answer exam questions written by generative AI and that I had to say something. One such perplexing misstep occurred when a student answered a prompt asking for a summary of Indian relocation in 1956 with an authoritative, robotic recounting of Irish relocation during the Civil War era. Another mistake occurred when multiple students independently listed one of the requirements of a speech reflection paper as “making eye contact,” which is essential for an actual presentation but comically incorrect for a submitted text document. I told my students that generative AI is essentially the average of all of the knowledge on the internet, and they needed to submit their own work that reflects the above average scholars that they are. Yet, as an academic committed to embodying and perpetuating lifelong learning, I know better than to dismiss any potentially helpful educational technology and so I immersed myself in the literature. My thinking on generative AI is evolving, but one thing that remains steadfast is that classrooms must remain a space for students to demonstrate their brilliance.

It’s important to consider the reasons that people use generative AI, as two in particular drive most of the conversation around its usage. The most common belief is that students harness the technology to locate answers they can copy and paste into their coursework. All of the flawed examples I referenced earlier employed bots to exploit this shortcut, but those students are not the ones who concern me most. The second group are those who use generative AI to appeal to their confirmation bias, meaning they ask the bots to find or even create sources to validate their beliefs. Referred to as a “hallucination,” AI can generate false, unsupported information that it presents with the same assurances as all of its other outputs. These missteps can be embarrassing for those who quote and cite them without validating their accuracy, but they have more sinister repercussions for users seeking reassurance for their limited or detrimental worldview. While these two motives for using AI technology are different, there’s little wonder that society rushes to their screens to seek answers without examining the repercussions.

Megan Garber’s Screen People: How We Entertained Ourselves Into a State of Emergency explores the ways people imagine themselves as an extension of our internet-based habits. Illuminating the steady progression from seeing oneself as separate from the technology to being intertwined with it, Garber conveys how “performance is so ingrained in American culture—as an aspiration, as an expectation, as a value—that it doubles as a script.” Noting that one of the highest compliments one can receive is that “your story should be a movie,” the text shows how having “main character energy,” and thereby making yourself the focal point of any interaction, can lead to isolation. Garber recognizes that screentime equates to “entertainment” that increasingly emphasizes a user’s biases in a reflective loop, and she offers both an astute means to critique where we are as a screen-centered society as well as the consequences of our collective refusal to disengage from it.

Despite our screen addictions, we must recognize both the negative repercussions of using generative AI and the reality that the technology is not going away. Although it took 90 years for one billion people to have electricity, 36 years for the same number to have the internet, and 16 years for smartphones to have as many users, it took only three years for one billion people to use generative AI. In 2026, international investment in the technology reached $725 billion, making it the biggest capital expenditure in human history. Beyond its effect on human knowledge, the impact this tech will have on the world in terms of fiscal and environmental costs is quantifiable yet unfathomable. Paradoxically, people are protesting against the creation of data centers that fuel generative AI’s computer servers while also rushing to use its products. In post-secondary education, in-coming students are looking to be taught the skills needed to meet employment demands, as faculty are balancing our AI concerns with our responsibility to students. One thing that resonates with me is that our graduates need a skillset to contend with the tech explosion, meaning we must educate them to work in the world they live in and not the one we may prefer. If generative AI alone can pass one’s class, then the instructor must rethink the pedagogy of their assessments before the tech is truly ubiquitous.

Josh Tyrangiel’s AI for Good: How Real People Are Using Artificial Intelligence to Fix Things That Matter explores the ways in which generative AI can advance innovation. Tyrangiel emphasizes that AI makes leaps not through mathematical prowess but rather through language models, underscoring how adept educators of all disciplines can use it to tailor both their assessments and student supports. AI is also being deployed to improve MRI heart scans, predict sepsis infections, and identify which households are putting contaminants into their recycling bins. Tyrangiel chronicles AI’s impact upon Operation Warp Speed, delivering millions of COVID-19 vaccines despite contending with numerous bureaucracies, as well as how it’s helping nonverbal people on the autism spectrum communicate their needs to caregivers. Tyrangiel’s overarching message is that “user feedback is the most powerful mechanism in tech,” and that if we want AI for good, we must use it as such.

Our students’ output exceeds the average of the internet, and we must give them an opportunity to exceed expectations. Although my thoughts on generative AI are evolving, I still have fears about its impact on humanity and the environment. Yet, I remain hopeful for what it can do to help manage natural resources and make governmental services more effective. We are facing an unparalleled disruption in the creation and perpetuation of knowledge, but one thing that I know for sure is that tribal college faculty will keep adapting and will thereby find new ways to ensure that our graduates can demonstrate their brilliance.

Ryan Winn, PhD, teaches in the Liberal Studies Department at College of Menominee Nation.

Editor’s note: The opinions expressed in the Writer’s Corner or any other opinion columns published by the Tribal College Journal (TCJ) do not necessarily reflect the opinions of TCJ or the American Indian Higher Education Consortium.

References

Garber, M. (2026). Screen People: How We Entertained Ourselves Into a State of Emergency. New York: Harper Collins.

Tyrangiel, J. (2026). AI for Good: How Real People Are Using Artificial Intelligence to Fix Things That Matter. New York: Simon and Schuster.