Straight From the Source: The State of AI Amongst Students

February 27th, 2024
WRITTEN BY HOPE HSIEH

Artificial intelligence is rapidly reshaping industries, and an important question is how the younger generations are learning to adapt to these shifts as they grow alongside the technology. As student ambassadors for The Browser Company, we conducted research to explore the nuances of how students—from high school to graduate school—regard and use AI in their daily lives. Through surveys and conversations, we gathered data and viewed trends on both the value AI can provide and the attitudes of the students.

The results were complex: students see AI as both a powerful tool and a potential crutch. Because of this, they worry about over-reliance on the technology, misinformation, and the potential for privacy risks.

Despite this, they also use AI daily, integrating usage into their academic, professional, personal workflows. Here's a closer look at what we found and why it matters.

tl;dr: STUDENTS SEE AI AS...

A powerful tool, and a potential crutch.

AI as a Tool, Not a Voice nor Collaborator

One of the most interesting takeaways was that students don't see AI use as an all-or-nothing proposition. Many don't feel the need to disclose the full breadth of their AI use—especially in cases of assistance—in their academic work. Not because they're being dishonest, but because they only use it for what they consider “menial” tasks: summarizing readings, proofreading drafts.

In the eyes of students, there's a common consensus that these aren't seen as acts of intellectual creation, but as clerical work that they can offload for AI to handle.

“Being able to organize and communicate your own thoughts in your own words is a valuable skill to refine, AI or not,” one student said. “As long as people aren't letting their critical thinking, speaking, writing, and other skills fall by the wayside [then that's fine].”

Being able to organize and communicate your own thoughts in your own words is a valuable skill to refine.

The Limits of AI-Generated Writing

AI is often used to troubleshoot code or refine writing. But when it comes to personal expression, crafting the actual argument of an essay or selecting ideas and making art, students want to maintain full control. “I use AI to brainstorm and sometimes workshop [my] writing, but I never take huge AI-generated chunks [because] it's usually not that good [or] doesn't sound like me,” a student says. Another agrees by clarifying that she doesn't use AI for full writing, just editing.

This hesitation and lack of full buy-in is widespread.

A key insight here was that many students cited using and preferring Grammarly for smaller fixes to their writing because other tools such as ChatGPT or Claude tend to rewrite too much and strip away the student’s natural style.

Because of this, students tend to use AI as a structural tool rather than a writing one. They might ask an assistant to generate an outline or to rephrase a specific line for clarity—but when it comes to fully formed sentences, they step in. Some will even rewrite the outputs entirely, and regard the AI’s response as a peer’s rough draft or inspiration rather than a final product.

This wariness of AI “voice” ties into a broader issue of trust: if AI can’t convincingly replicate human nuance in writing, how can students trust it for more complex tasks? Such skepticism plays a significant role in why students have such strict limits on AI’s involvement in both their creative and intellectual work.

The Fear of Skill Atrophy

A recurring theme that appeared throughout the research was a concern that AI is making students less willing or able to engage deeply with different types of academic and societal material.

“I’m concerned about my over-reliance on AI, I want to keep my critical thinking sharp and I fear that using too much of it in my workflow will cause my skills to atrophy.” One student voiced more concern over soft skills whilst another fretted over the degradation of more basic ones such as arithmetic and algebra.

A third student took a more definitive stance by saying she does not and will never pay for anything AI-related because it is already difficult to develop her critical thinking skills. If she begins to rely on AI, she won’t pay to “get stupider.”

The instant availability of AI-generated responses that students value about the technology is a double-edged sword as it makes it tempting to bypass the steps and effort of working through complex problems, leading to a growing fear of an erosion of independent thought. 

Also in student minds is the broader concern over learned habits. One student touched on expectations: “[Large Learning Models] spit out answers within milliseconds, which is impressive, but this can decrease our attention span in the long run.” In regard to this conversation, there is less consensus. Some other students argue that AI can be used to enhance, rather than replace, understanding.

One particular student dissected their personal interactions with AI, saying that “[When using AI] I examine the process it is using so that I understand and comprehend how to do the problem, rather than just copying the output…I aim to use it in a way that still educates me on the subject at hand.”

Future Uncertainties and the Value of [Human] Work

Extrapolating their fears outwards and long-term, many students expressed unease about the prospects of the future job market as AI automates more and more tasks.

Some were broad: “I worry humans will forget how to do human things and we’ll lose our culture”.

Others were more specific: “What scares me the most is [AI’s] capacity to eventually replace humans in the workplace and in other facilities”. In tandem with this fear, held in balanced conflict with the desire to have AI be efficient, is the fear of skills being devalued with quotes such as “I think AI shrinks the gap between the capable and less capable” and “The value of hard work will be lost because many things will become easy [for everyone].” 

I worry humans will forget how to do human things and we'll lose our culture.

And the roles students are most fearful of becoming integrated with AI? The creative ones.

Many students were adamant that AI-generated art and music lacked depth and resonance— because they believe the human aspect is yet to be fully replicated by the technology and also because they believe the aforementioned humanity is what makes art “beautiful and valuable.”

The theme appears to be that critical thought and creativity are traits associated with being a human and even the possibility of losing that control to AI strikes deep fear with most students.

Privacy and Safety, Trust and Misinformation

Finally, as with the larger society, students have the same fears as most do about emerging tech, primarily in regards to privacy and transparency. Many students expressed discomfort and subsequent hesitancy to share personal data with AI companies, some stemming from the recent virality of AI and some from the current social and geo-political influences.

“It’s scary to think about how [companies] handle [your information], especially since it’s not really regulated,” one student said. For the most part, this fear surrounding privacy emerges from a distrust that is coupled with misinformation: as students noted that the research insights that LLMs might pull as responses can contain errors that seem simple and yet are indicative of a larger gap to stability.

It's scary to think about how companies handle your information, since it's not really regulated.

One student did caveat this margin of error by saying first that AI “doesn’t always have the most accurate information and I have to always check if the output is right which takes away from using AI in the first place. I still learn this way so maybe it’s a great thing!”

Wrapping Things Up:
How Students Actually Use AI Today

Despite these concerns, AI is deeply intertwined with student routines. In summary, here’s how they use it: summarizing and explaining concepts for school, solving and debugging specific errors in work and coding instances, organization and as a sounding board for personal uses, and sometimes as a replacement for Google for quick queries and answers. Interestingly, those who use AI more tend to be more adept at steering and tailoring the technology through advanced prompting strategies. While other students, more wary of limitations and weaknesses, use it sparingly and yield more variable results. 

Popular tools & Frequency of use

95.3% of our participants have used ai. The most popular tool of choice? chatgpt. >50% reported using at least one additional tool beyond mainstream options.

Ultimately, students are neither blindly optimistic nor entirely dismissive of AI. They see potential, and effectiveness in regards to enhanced learning and productivity, but they are also keenly aware of the pitfalls.

Many have drawn an unspoken boundary: AI is for repetitive and tedious grunt work, while meaningful thinking is best left to humans.

This nuanced approach suggests that the future of AI in education and for students isn't about banning or embracing it wholesale. Students believe that change is coming for the schooling system, whether by choice or by force of the tides, and so the focus ought to be on how to teach discernment surrounding AI.

If students continue to view AI as a supplement rather than a substitute for their own intellect, it could develop into a powerful asset without diminishing the human ability to think critically and create meaningfully.