Reading Between The Lines : NLPs, Critical Comprehension, and the Way We Teach
When one thinks about AI in the current zeitgeist, they are likely most drawn toward NLPs (natural language processing systems) and their effects on the world. The discussion around such AI models has overwhelmingly been around how they bypass the education system by removing many of the learning experiences that come with writing essays by hand. With the speed at which NLPs are improving, one could reasonably claim that there will soon be a day when human and machine-generated writing becomes nearly indistinguishable.
This then begs the question of the utility of refining one’s essay/general writing skills beyond knowing how to prompt and tweak an NLP’s output. The fact that many assignments can be completed with little to no human input shows that perhaps there are flaws in how we teach writing in the North American education system.
Here, I’d like to show why the impact of NLPs on teaching from grade school to early tertiary education may increase critical comprehension and general education quality. This translates to potentially increasing reading ability/level (still an issue in North America), political discussion, work efficiency and preparedness for students entering the workforce, the creation of new job fields, and research paper coherency.
How could AI increase critical comprehension in future generations if appropriately applied?
In a way, AI points out a flaw in how we teach writing in late high school: often, there is minimal critical comprehension occurring. This was the case even before the recent boom in usage and popularity of NLPs.
For example, if asked a somewhat common and shallow question about the symbolism of some element in Hamlet, a slightly clever and lazy student could find a perfectly coherent and potentially superior response using Google. ChatGPT only highlights this issue by removing the step of copying down the response and changing a couple elements.
In both cases, our instance of a smug student using Google, ChatGPT, or any other online tool to help write their essay has minimal opportunities to think critically.
I believe all this fanfare can be positive as it forces education systems to become more specific and intentional when designing learning assignments to prevent accurate responses from NLPs. Thus, we kill two birds with one stone and reduce the impact that easily accessible information and help have on students.
What critical comprehension/greater reading comprehension does for the world and why it’s important.
Forcing students to think critically also has many other far-reaching benefits. Indeed, many North Americans still have subpar reading comprehension, as 54% of Americans currently read below an equivalent sixth-grade level. This leads to issues when it comes to understanding complex topics, such as the ones commonly found in local and international politics. A greater average of political knowledge strengthens democracy; citizens are more able to hold leaders accountable and are less susceptible to misinformation. Increased political understanding may also allow for a robust civic culture wherein individuals understand and act upon their responsibilities as participants in the democratic system. Literary comprehension also benefits the STEM field by improving human understanding of data and allowing for better-written papers. Because much of the brain’s development and, thus, thinking patterns are established in the teenage years to early adulthood, it is essential to teach people how to think critically at a younger age.
The Place of AI in Education and the Future
I am not asserting that students should forgo learning how to write an essay altogether. AI is a tool just like the calculator; although we all carry one in our pockets, it is still essential to understand the fundamentals to build upon them, whether it’s multiplication tables or the elements of a synthesis essay.
I believe that writing is one of those courses that are often pushed aside as they are more complex (or impossible) to grade thoroughly and objectively. Thus, there can often be redundancies in how courses are taught and an overly strong focus on those smaller elements that are easier to quantify (structure, grammar, et cetera).
An interesting new approach, process writing, is provided by the British Council. This system prioritizes active feedback throughout each step of a semi-structured process. Initially proposed by White and Ardnt in 1991, the authors claim that the current focus on grammatical minutiae “improves neither grammatical accuracy nor writing fluency.” Though more labour-intensive, teaching this way could allow students to develop their own methods to break down writing into smaller tasks and other tasks in their daily lives.
Though I am no educator, I believe there is some sense in allowing for the complete eschewing of the actual writing in essay form in the later years of a student’s education and instead focusing on formats such as prompts and bullet points that show one’s reasoning. Some may forget how an essay is shaped, but can the same not be said for long division or the other redundant mathematical tools we’ve been taught?
These days, most future-thinking educators agree that AIEd (Artificial Intelligence in Education) will have some concrete place in the education system. However, what AI’s place is has yet to be hammered out. According to a paper by Holmes et al. from UCL, “AIEd is not about replacing teachers or reducing their role in education. Rather, it is about providing teachers and learners with the tools they need to succeed in an increasingly complex and interconnected world.” It is notable that nearly the same thoughts were voiced during the popularization of the internet.
To add to that sentiment, we look at a quote from Andreja Starčič, a professor at the University of Primorska in Slovenia. “AI is not a substitute for human intelligence, but a complement. AI can augment human capabilities and enhance learning outcomes, but it cannot replace the human touch, the social and emotional aspects of learning, and the creativity and critical thinking skills that are essential for lifelong learning.”
For now, we are left to ponder whether AI should be a new subject, a direct mediator, or even a supplementary assistant. This is doubly the case when we look at the poor technology integrations highlighted by the COVID-era in the past decade. If we cannot yet adapt to an online world, are we ready for an artificially intelligent one?
Only a couple decades ago, people found it a great loss how the average person’s mental math has degraded over time. But the thing is, complex mental math is no longer all as significant. The same might be applied to fundamental essay writing. We could maybe even focus on writing prompts where we provide the requisite data and information to an NLP.
I am currently researching this in my QMind project QM-1. Timnit Gebru, a highly influential researcher in the field of AI ethics, recommends that future data processing be done by trained professionals, which would improve NLP quality and help prevent outsourcing to and the potential exploitation of the global south. Teaching students how to provide information to NLPs to get specific desired outputs could be a stepping stone to this new potential form of gainful and critical employment. It would also save time writing when they could instead be researching or actively working on their project.
Conclusion
To summarize, the rise of tools that significantly speed up the writing process could, if used with a solid fundamental understanding of writing and reading comprehension, lead to enormous strides in discussions throughout the general public, more significant work efficiency, and generally forces changes to our education system which would target literary understanding instead of an understanding of literary rules. Like calculators, I believe that the skill of essay writing will similarly fall by the wayside. If such is the case, one may ponder what other elements in our education deemed “essential” could be mostly automated. Perhaps there will be a shift towards developing our soft skills such as critical thinking, adaptability, communication, and emotional intelligence. Or maybe there will be a far stronger focus on mental and physical health. Whatever may occur, the marching progress of technology has shown us that it is far more essential to see what we understand than how much we know.
This article was written for QMIND — Canada’s largest undergraduate community for leaders in disruptive technology.
