Whether it’s to summarize a jargon-heavy excerpt for class or to outline reasoning in an argumentative essay, there exists an unspoken shame in using AI among those who consider themselves academics. This is understandable. The fundamental pursuits of academic life, such as intellectual freedom and truth, seem diametrically opposed to the ends of the tech industry and their AI products: growth for the sake of growth.
However, the question at hand is not whether AI is an intrinsic good for humanity, but rather a discussion of whether it should be adopted as a learning tool by Tufts. Refusing to adapt to our changing world in the name of academic purism risks misalignment with the Tufts mission statement of “creation and application of knowledge.” In an age when the value of a college education has been repeatedly contested, it is crucial that Tufts prepares its student body for the inevitable integration of AI in all aspects of life, particularly in employment.
While the word “inevitable” might seem sensationalist, there has been ample proof that AI will become a central part of the workforce. Checking the headlines on any given day proves just how much the economy is revolving around AI and the tech sector (which one might argue have become synonymous). Pinterest attributed a 15% reduction in the workforce to AI restructuring, Amazon is slashing another 16,000 jobs for “organizational change,” and the national hiring rate today is even lower than pre-pandemic levels. We may be in the trough of the J-curve now, and thereby change may not be apparent, but it is indeed underway. Jaw-dropping investments into the tech sector will guarantee that, continuing to encourage the synthesis between AI and the rest of our lives.
Most AI tools that students are familiar with are Large Language Models (LLMs), tools that are designed to understand and generate text above all else. There are, of course, specialized AI tools such as Claudebot or Github Copilot that are more prevalent in STEM fields, but the nature of the study of STEM makes the integration of AI more generally accepted. The specific argument here is that if STEM subjects are the only ones to adapt, the humanities will get left behind in the ever-changing landscape of academia.
A 2025 University of Southern California study found that when students used LLMs with no guidance from professors, they were more likely to utilize them as a shortcut to obtain quick answers, whereas understanding of concepts deepened and critical thinking increased when training and specific instructions were provided. A Massachusetts Institute of Technology study conducted in 2025 also supported these findings when it tested four groups, who used four different configurations of tools in an essay writing task: brain-only, search engine only, LLM-only, and brain-first-to-LLM. As expected, brain-only users exhibited the highest level of ownership and neural connectivity, and LLM-only users showed the lowest. However, the group that used their brain first to outline and ideate and then supported their work with LLMs surprisingly showed the highest levels of memory recall and activation of occipito-parietal and prefrontal areas of the brain. Repeated studies have proven that AI is indeed cognitively catastrophic when used without regulation and intention, but the correct training and information can make it just as useful as a search engine.
The capabilities of AI to “generate” ideas or “understand” concepts can seem impressive at first glance, but there are clear benchmarks that AI has not surpassed yet in terms of intelligence compared to humans. One must have an accurate understanding of its strengths and weaknesses to transform it into an efficient learning tool. In the same way we are provided with media and digital literacy programs, it is crucial that we grasp the true nature of AI so we can mold it to our needs instead of being molded to the world it creates.
There are already talks of how AI is going to deepen, if not create, a new class divide. The Guardian reported that many private elementary schools are incorporating AI-centered courses that deepen the understanding of the algorithm itself, and at the DAVOS economic forum, CEO of Blackrock Larry Fink, stated that AI could become capitalism’s next big failure after 30 years of its inability to deliver equal benefits across socioeconomic class.
AI in education is, in essence, not the attractive option but the necessary one. In a society that values the final product over process and productivity over all, AI integration is inevitable and we must adapt if we want any part in the future.
Artwork by Grok AI