What Does AI Mean for the Future of Education?

The arrival of Open AI’s GPT-4 crossed a threshold that captured the imagination—and simultaneously aroused the fears—of millions, making it clear that the world would never be the same. Promises of General Artificial Intelligence, which had seemed many decades away, suddenly felt within reach.

While GPT-4 and other AI models like it are not without flaws, they are good enough in their nascent form to begin transforming many fields, including the field of education, where they’ve already had a profound impact.

Why We Should All be Paying Attention

We are only in the first chapter of this story. It will be some time before Large Language Models (LLMs) like GPT-4 and Anthropic’s Claude give way to the next generation of Generative AI systems, but the arms race is on—and the funding is pouring in to develop this technology to its full potential.

Artificial Intelligence is already all around us: our smart thermostats, our turn-by-turn driving directions, spelling and grammar checks, TV and music recommendations. We have already accepted a world where humans are subordinate to machines in the narrow domains of Chess, Jeopardy, Go, and Protein Folding.

But GPT-4 showed us a world where humans could have a thought partner and teacher in myriad domains, an intelligence that could rock the LSAT, the Bar Exam, the GRE, and a variety of AP exams. An intelligence that could impersonate Shakespeare, plan your next vacation, give decent parenting advice, solve your chemistry assignment, and paint like Van Gogh. If you’re not impressed, you’re not paying attention.

A Direct Impact in Education

Students are using GPT-4 and other LLMs like it to expand their knowledge, complete assignments, generate ideas, and proof papers. In a recent webinar on AI in education, hosted by Jeff Selingo, Ben Breen of UC Santa Cruz spoke of his peers in the Spanish department noticing that suddenly all the Spanish essays were being turned in without errors: no comma errors, no grammar errors. Lilach Mollick of UPenn’s Wharton Interactive noted that any homework assignment—for a student who is good at prompting (skillfully entering data into GPT-4 to optimize and tailor the output)—could be done using AI.  

If an AI can complete any assignment, educators are going to have to begin to rethink both their pedagogy and how they assess student knowledge. Breen noted that given the advent of GPT-4, essay assignments are no longer a real signal for what a student knows/is able to do. Teachers are going to need time to catch up to the new reality, as the students are so far ahead. Breen noted a range of faculty responses to the new developments in generative AI: some of his colleagues were excited, others were worried, while others were completely oblivious.

Some academics are primarily concerned about the impact of AI on cheating, but Breen argues that this is a limited perspective. Breen, an Associate Professor of History, argues that while cheating is happening, that’s not the story here. “When we write about this in 100 years, we won’t be talking about cheating.”

Students are likewise embracing this technology at a quick clip. One survey found that 30% of college students used Chat GPT for school work in the 2022-2023 academic year. The adoption is only going to increase across the student population.

Potential Applications for College Admissions Offices

While students are seeing AI’s potential to synthesize information, to generate or improve written assignments, to translate documents, and other applications, college administrators and admissions offices are seeing the potential of the technology to customize and personalize outreach messaging and marketing, to read and score transcripts and applications, and to synthesize and organize and prioritize mountains of data.

Slate, a leader in CRM for college admissions with 1700 member schools, announced at its 2023 Slate Summit new AI-powered features including the ability to summarize student essays and recommendation letters (i.e., the Slate Pre-Reader), calculate GPAs, and expedite transcript review.

There’s a world that is not so hard to imagine where one AI identifies the students to recruit, another AI writes the application for the student, and another AI reads and scores the application. This is a story about humans gaining efficiencies, but losing some of that human contact. Is that a price we are willing to pay?

Either way, colleges are already jumping in to use this technology. One recent survey found that 80% of colleges will be using AI in admissions by 2024.  

What Will These Learning Machines Teach Us?

AIs will give us mountains, petabytes of data, at our fingertips, and will force us to become good curators and editors, but as Charles Isbell, Provost at UW Madison has asked, once we embrace this technology, will we remain good writers? We already allow machines to handle the spelling, and the grammar. Will the next step be to offload the entire task of writing to AI?

We tend to get excited by the potential efficiencies brought by new technologies. But every technological advancement comes with some cost. Has spell-check made us better at spelling? Google Maps made us better at orienting ourselves? Calculators made us better at mental math? In most cases, no. We sacrifice something, some capacity, to gain efficiency.

On the other hand, technologists understand that this technology is transformational. In the way the personal computer, the internet, the cell phone influenced our daily lives, this technology will become similarly embedded in our daily lives. While the transformation is upon us, it will take some time for the broad adoption of this technology. Iterations upon iterations and improvements upon improvements will leave us with AIs that we can today only imagine. It’s an exciting time.

Like many companies, Applerouth is actively examining all the ways this new technology can assist us, our students, and their families. The power to analyze, synthesize, and personalize information and data is very real. There are so many use cases for generative AI, and we hope to harness this new technology to improve the experience of our students and enhance their academic outcomes.