Key points:
- AI can replicate knowledge transfer quite well, but it cannot perform the human act of teaching
- Students need more than AI access–they need AI rights
- Colleges are adopting AI faster than they can govern it
- For more on higher education and AI, visit eCN’s AI in Education hub
AI’s rapid integration into the higher-education landscape has prompted a period of profound structural reassessment. For decades, new technology adoption in education has often been driven by a “technology for technology’s sake” mentality–a pursuit of digital innovation that typically prioritizes the novelty of the tool over the efficacy of the intended outcome.
However, the current AI inflection point demands a more disciplined strategic shift. Rather than viewing AI as a solution searching for a problem, higher education must pivot toward using AI technology tools to solve fundamental, long-standing learning issues. The strategic priority is no longer the mere implementation of digital platforms, but the intentional design of environments that facilitate deep, measurable cognitive development. This technological surge has exposed a significant structural gap within the profession: The critical distinction between subject matter expertise and teaching/learning expertise.
AI has acted as a catalyst, exposing this pedagogy gap by demonstrating that while machines can replicate the transfer of information with startling efficiency, they cannot replace the structured design required for true education. This realization serves as a vital turning point for the profession. It forces a move away from the raw transfer of knowledge as a primary value proposition and toward a more sophisticated understanding of the science of learning. This shift reveals that the most pressing challenge facing higher education is not the AI itself, but the underlying pedagogy gap that has been masked by traditional instructional models for generations.
Analyzing the pedagogy gap
Historically, the higher-ed academic profession has operated under the assumption that deep content expertise is a sufficient proxy for teaching ability. In the hierarchy of research-heavy institutions, pedagogical training, defined as the formal study of how people acquire, retain, and apply knowledge, has often been treated as secondary, if it is addressed at all. Professors are frequently recruited and evaluated based on their ability to conduct original research and publish in peer-reviewed journals, with teaching responsibilities viewed as a secondary or even tertiary endeavor. Consequently, many instructors enter the classroom with an immense mastery of their specific discipline but possessing few, if any, basic teaching skills. This lack of foundational pedagogical knowledge creates a significant barrier: It is difficult to build new, AI-enhanced learning models when the foundational understanding of how students learn is largely absent.
There is a stark disparity between explaining subject matter and teaching. While content knowledge remains the bedrock of academia, it is no longer the primary differentiator in an environment where AI can summarize complex theories in seconds. In this new landscape, content must be supported by delivery techniques and cognitive learning theories.
A pervasive issue in this context is the tough professor fallacy. Often, instructors who are hard graders or exceptionally difficult are misidentified as being rigorous and therefore, by default, effective. Being difficult is not synonymous with being an effective educator. This reputation often masks a lack of instructional understanding, where the burden of learning is placed entirely on the student’s ability to decipher the instructor’s expertise rather than the instructor’s ability to facilitate mastery.
Without an understanding of instructional intervention, faculty risk using AI merely to make the act of teaching easier or to fill instructional time, failing to address the actual cognitive issues students face. This disconnect underscores the urgent need for faculty to develop instructional skills that bridge the gap between knowing a subject and effectively teaching it.
AI as a precision tool
To understand the role of AI in the modern classroom, we must view it as a precision tool analogous to a scalpel. The efficacy of a scalpel is entirely dependent upon the skill, intent, and specific input of the user. In the hands of a trained professional, it is a powerful instrument for improvement. But in the hands of those without the concomitant expertise, its utility is lost, or worse, it becomes a source of harm. AI can be used as an instructional tool, but ideally only when combined with an understanding of how to address learning issues.
Some believe students should freely experiment with AI, while others completely ban it. The rapid rise of AI has left higher education unsettled. Academic researchers argue that both approaches are ineffective; formal instruction needs clear goals and measurable objectives. AI should address specific learning challenges, not replace structured curricula. Because AI’s integration into society is inevitable, outright bans are not practical.
Education is shifting away from the lecture-then-test model, where instruction is evaluated and measured primarily by contact hours. In AI-integrated settings, instructors’ value depends on student learning outcomes. They must become designers of learning experiences, using AI to target specific gaps. Focusing on clear outcomes ensures technology supports precise educational goals. Course delivery will require instructors to develop new skills, and although integrating AI is rapid and challenging, staff need guidance and training.
AI intervention
The integration of AI into the curriculum cannot succeed if it is treated as a standalone feature or a technological add-on. Instead, it must be supported by established theoretical scaffolding to ensure that students are not left to self-teach without a map. There is no empirical evidence to suggest that most students can effectively self-teach complex subjects without guided intervention.
By applying learning theories such as those proposed long ago by Lev Vygotsky, we can create different frameworks for exactly when and how AI should be integrated. One of the greatest challenges of current AI implementation is that many instructors are “starting in the middle,” focusing on the tool itself before understanding the cognitive process. We must instead start at the beginning with at least some basic theory of how learning occurs, and what type of learning methods might work best given specific content.
The concept of the Zone of Proximal Development (ZPD) and the practice of scaffolding, among many other learning theories, help to provide insight. To apply the idea of ZPD, we must categorize student capability into three distinct levels:
- Tasks the student can do without assistance: At this level, the student has already achieved mastery in a specific topic with no further instruction necessary.
- Tasks the student can do with assistance: This is the desired zone or sweet spot. This is where AI-assisted instructional intervention is most powerful. AI tools can provide real-time feedback loops, scaffolded hints, and individualized pacing that bridge the gap between a student’s current ability and mastery. Most importantly, AI is a 24/7 assistant which promotes continuous learning without pauses for instructor assistance. Students tend to create work nights/weekends with limited or no access to instructional support.
- Tasks that are impossible even with assistance: At this level, the material is beyond the student’s current cognitive reach. Attempting to use AI to force learning here can lead to frustration and a lack of retention.
As AI becomes increasingly proficient at the raw transfer of knowledge, this aspect of education is rapidly becoming a commoditized resource. If a student can input a list of learning objectives into an AI tool and receive a clear, accurate explanation of the material, the traditional role of the teacher as a content relay becomes effectively obsolete.
What remains uniquely human, and therefore more valuable than ever, is the ability to provide the extrinsic effect: the motivation, support, and encouragement that sustain a learner’s progress.
The failure of Massive Open Online Courses (MOOCs) serves as a critical cautionary tale. Despite offering high-quality content for free from prestigious universities such as Stanford, MIT, and Harvard, they have suffered from abysmal completion rates because of student isolation and the lack of human connection.
The distinction between knowledge transfer and teaching is also relevant in the locus of control of students. Most learners require some level of extrinsic support to maintain focus and persist through difficult material. This need varies across demographics:
- Children: Possess a low internal locus of control and require constant extrinsic motivation and guidance.
- Adults: Often have a stronger internal locus of control and may find autonomous, AI-driven learning highly viable.
- College students: Typically fall in the middle. While moving toward autonomy, many still require significant human connection and encouragement to remain engaged. The locus of control of college students is highly variable.
This reality introduces an economic analogy for the future of education: the Airline Tier Model (ATM). In this airline-type model influenced by AI, every student in a course is seeking the same destination (the learning outcome). However, the level of support they require can be categorized somewhat as pricing for airline seating. Some students in a course may only need “Economy” service–minimal human interaction and primarily AI-driven support. Many students are independent and autonomous learners. We often see this by students who rarely attend class or have low online participation yet succeed in the course. Other learners require “Business” or “First Class” levels of service, involving high-touch human interaction and personalized instructional support. Importantly, some students move between the service levels throughout a single course.
AI could provide the viability to offer these different levels of service, but the human teacher remains the essential differentiator. While the world may be evolving toward more autonomous learners, the human connection remains the catalyst that prevents the isolation seen in many failed digital models.
Institutional and instructional evolution
Schools are essentially facing a strategic necessity to evolve from content repositories and conduits into centers of learning excellence. As learning materials have increasingly become commoditized, a degree is not evidence of access to information, but rather as a certification of demonstrated competence. Consequently, and perhaps surprisingly, teaching practice will become the primary differentiator for institutions in this new age of AI. Schools that can demonstrate superior student success through high-quality instructional design and human-centric faculty engagement will thrive, while those that rely solely on the prestige of their content could be superseded by more accessible, cost effective, AI-driven alternatives. There is little doubt that new alternative degree/certificate providers will emerge with AI mediated courses and degree programs. Sal Khan recently announced such a vision to rival ivy-league level educational quality and fraction of the cost using AI. Of course, that is a big idea and lofty goal, but clearly there are seeds of innovation at our feet. Whether this idea is a bubble or behemoth remains to be seen once there is some empirical research.
As AI continues to commoditize knowledge, the attraction for students will be how faculty act as a catalyst for success through pedagogical expertise and human interaction. Ultimately, the successful integration of AI into higher education requires us to stop starting in the middle by focusing on the AI technology itself. We must start at the beginning with a much deeper and broader understanding about how people learn. AI can replicate the knowledge transfer quite well, but it cannot perform the human act of teaching. By addressing the pedagogy gap and prioritizing the science of instruction, higher education can redefine its role and ensure that technology serves as a tool for genuine, human-led advancement. Only then can we bridge the gap between knowing a subject and truly teaching it.
Dr. Mark Taormino is a professor in the Computing & Information Technology Department at the College of Southern Nevada. His research interests are focused on Instructional Technology and using technology as an instructional and learning catalyst. In addition to teaching computer and technology courses, he works closely with faculty to integrate pedagogy and technology into teaching practice.