Key points:
- When designed as a scaffold, AI expands learning
- Building the AI-ready graduate
- Recentering discovery as the core of learning through AI
- For more news on AI’s impact on learning, visit eCN’s AI in Education hub
Meaningful learning should not be about memorization or developing the ability to solve problems for which solutions are already known. Rather, it should focus on doing, experiencing, questioning and deciding, engaging with ideas in context, testing them against reality, and developing the ability to act thoughtfully under conditions of uncertainty. This is the foundation of both a Socratic and a polytechnic education, where the emphasis is not simply on the acquisition of knowledge, but on critical thinking, challenging assumptions, and developing the capacity to use knowledge adaptively, critically, and responsibly.
Over time, higher education has drifted from this foundation, evolving to an environment more aligned with an assembly line where learning is reduced to the structured delivery of content and to an assessment methodology that rewards the repetitious solving of known problems rather than developing the ability to frame and approach new ones.
Framed primarily as a tool for access, acceleration, or support, AI is often positioned as a means of arriving at answers more efficiently. While such capabilities are valuable, they risk reinforcing the very limitations that higher education must work to overcome. The central failure is not the lack of use of AI in higher education, but that we use it to reinforce a fundamentally outdated model of learning.
The challenge before us is not simply to integrate AI into existing models, nor to use it to improve efficiency within current structures. It is to reconsider what we expect learning to achieve and to design environments in which students engage with complexity, explore “what-if” scenarios that test assumptions and develop critical thinking skills, and develop judgment through experience.
A twentieth century education, without appropriate modification, is neither relevant, nor advantageous, in a twenty-first century economy. The question is not what we teach and what students learn, but whether we are preparing them adequately for a future of constant and rapid change. A student who can give a detailed explanation of the laws of thermodynamics or who can discuss the great books at length, but who has no understanding of logits, transformers, or how GenAI could affect the security of our information networks and beyond, is not well prepared for success. Addressing this is critical to maintaining, and increasing, the relevance of higher education. AI, when used thoughtfully and intentionally, offers a powerful mechanism to enable precisely this, serving as a scaffold in enabling deeper engagement, broader exploration, and iterative individualized learning at scale.
When used as a scaffold, AI allows learners to move beyond solving problems to interrogating them, discovering not only what works, but why, under what conditions a feasible solution may fail, and the larger societal context of those changes. In doing so, it enables forms of iterative, experience-based learning that have long been central to engineering and polytechnic education, but that can now be extended more broadly and more consistently. This is a transformational shift from learning as the accumulation of knowledge to learning as the development of capability.
From tool to scaffold
When used without intention, AI can indeed collapse learning into the expedited retrieval of answers. However, when used with intentional design, it can expand learning into a process of exploration, evaluation, and synthesis. In this format, the scaffold is not defined by the presence of technology, but by the structure of the engagement enabled. Students begin by working directly with material, developing foundational understanding through interaction with learned instructors and through their own efforts with data, models, and concepts.
As these skills are developed, AI can act as a “study-buddy,” critic, and collaborator, assisting in challenging assumptions, proposing alternatives, raising questions, and opening new avenues for inquiry. In this stage, the learner remains central to the process, evaluating outputs, determining their validity, and deciding how to incorporate or refine them. As this interaction evolves, students begin to develop perspective not only on focus of study, but on their own reasoning processes. They identify gaps, pursue new lines of inquiry, and refine their understanding through iteration.
Mastery is thus developed through repetition, practice in an ever-widening sphere of understanding, pressure-testing not just the topic of focus but the larger context and then moving to prototyping either through extensive simulation or through actual development.
Ultimately, the process enhances the learner’s ability to integrate, synthesize, and construct solutions that are more robust and more thoughtfully developed. This incorporation of AI as the scaffold that enables this evolution effectively transforms largely passive consumption of information present in higher education today to an active, iterative, and reflective mode of learning.
Learning in context: From problems to judgment
The movement from memorization and dualistic thinking that suggests and rewards the single “correct” answer to in-depth critical thinking and synthesis is perhaps most apparent in disciplines such as engineering, where problems are shaped by assumptions, boundary conditions, competing objectives, varying degrees of uncertainty, and the overall consequences of decisions, rather than a single answer. A calculation may be entirely correct, and yet the resulting solution may still be inappropriate, inefficient, or even unsafe when considered in context.
It is within this distinction, between correctness and appropriateness, that deeper learning occurs. A scaffolded approach shapes engagement with such problems, shifting the emphasis from the current focus on a single solution to interrogating assumptions, exploring alternatives, testing how solutions respond to changing conditions, and identifying potential limitations and unintended consequences. The ability to engage with complexity, question assumptions, and make informed decisions in the absence of certainty is the core of true learning. While this is possible through deep and extensive interaction between learners and dedicated, highly qualified faculty at the very best small private institutions with extremely low faculty-to-student ratios, this is impossible at scale.
AI, when used as a scaffold, can fill this gap, expanding the learner’s capacity to engage in the process of deep exploration while simultaneously enabling more rapid iteration, deeper reflection, and more pertinent integration of knowledge with context. It results in the inculcation of critical thinking by learners that has largely been absent in higher education today.
Building a pedagogical advantage
A pedagogical model built on this foundation prioritizes the development of interrelated capabilities that extend beyond content mastery. Students first develop a strong foundational knowledge, which is then extended through the cultivation of judgment and decision-making capacity, particularly in situations characterized by ambiguity and competing constraints.
Equally important is the development of informed skepticism leading to the recognition that outputs, whether generated by AI or derived through traditional methods, must be evaluated, validated, and, where necessary, challenged. The ability to engage in creative problem-solving, explore alternatives, and navigate uncertainty requires the ability to interpret and synthesize information, integrating disparate inputs into coherent and defensible understanding. In a scaffolded approach, learners develop a systems-level perspective, building an appreciation of interdependencies, broader implications, and the cascading effects of decisions made within complex environments, thereby providing critical experiences that are essential for success in the workplace.
We must recognize that in a world where the generation of acceptable answers and polished reports can increasingly be generated by AI, the differentiator is no longer what graduates know, but how they think, frame problems, evaluate solutions and make decisions when information is incomplete and the constraints are ambiguous. Institutions that continue to emphasize the dualistic form of learning risk a growing disconnect between education and practice and hence a rapid decline into irrelevance. In contrast, those that intentionally cultivate contextual critical thinking and the ability to engage with complexity through the “what if” and “why” will define the model of higher education that maintains and strengthens relevance and value for the future.
A return to first principles
It is understandable that the integration of AI into learning raises concerns regarding the potential erosion of basic skills and knowledge. Such concerns, however, reflect a misinterpretation of what a scaffolded model represents. In the tradition of Aristotle’s Lyceum, education was not conceived as the passive acquisition of information, nor as the narrow development of discrete skills, but is positioned as the cultivation of reasoning, the development of judgment, and the pursuit of wisdom, through the development of abilities to act thoughtfully and effectively within complex and uncertain environments. A scaffolded AI model for learning aligns directly with this progression.
Through engagement with AI as a critic and collaborator, students are required to reason, to interpret outputs, examine assumptions, and understand underlying principles. Through the evaluation of alternatives and the assessment of trade-offs, they develop judgment. Through repeated cycles of reflection, synthesis, and application, they begin to approach wisdom.
Rather than producing workers at the expense of thinkers, which is often the argument against both the use of AI and the expansion of the polytechnic model, the scaffolded model demands increasingly higher levels of thinking.
Redefining learning, teaching, and the institution
The adoption of a scaffolded approach necessarily requires reconsideration of institutional structures. Current static curricula must evolve from fixed sequences of content delivery to dynamic, integrated pathways for exploration, application, and even learning through failure, with the scaffold enabling rapid recovery and forward progress. Traditional barriers, such as rigid prerequisites and progression defined within fixed time frames, must be reexamined in favor of models that allow flexibility in both path and in the method of attaining competency while supporting the progressive deepening of knowledge. The role of the faculty must evolve from content provider to expert guide, designer of transdisciplinary learning environments, and facilitator of inquiry and the navigation of complexity. This transition requires intentional design, institutional realignment, and a willingness to re-envision long-held beliefs about structure and functionality. Without these, AI will just become another technology reinforcing current limitations rather than enabling a re-envisioning of learning.
At the institutional level, this represents a shift from gatekeeper of knowledge to enabler of learning, supporting designed educational experiences aligned with the realities students will encounter beyond the academy.
A call to action: Designing for scaffolded learning
If AI is to function as a true scaffold for learning, institutions must move beyond bolting new tools into existing frameworks and shift from a system of discrete assignments to specially designed sequences that intentionally move students through stages of understanding, critique, exploration, and synthesis. The emphasis must shift from the dualistic mode of correct/incorrect to the evaluation of reasoning, requiring students to articulate not only what they conclude, but how and why their thinking has evolved along specific hypotheses. In this mode, AI is not a substitute for effort, but rather a structured support that enhances the challenging of assumptions, generation of alternatives, and exploration of unknowns. This will require substantial investment in faculty, making their role even more essential and critical to the learning process. Graduates of such environments will not only know more and be able to access more with discernment, but will be better prepared to act, to adapt, and to contribute meaningfully to society, in professional practice, research, and innovation.
The question is no longer whether AI will shape the future of learning. It already is doing just that. The more consequential question is whether it will merely be used to accelerate access to information or be used intentionally designed as a scaffold. When used as an isolated tool, AI risks reinforcing the very limitations that have eroded the relevance of higher education. When designed as a scaffold, it expands learning into a process of enquiry, discovery, iteration, and judgment, enabling the key capabilities that cannot be automated and should not be outsourced.
Institutions that recognize and act on this distinction will adapt and redefine the scope and relevance of a twenty-first century education. Those that do not risk producing graduates proficient in generating answers to known questions, yet who are insufficiently prepared to the realities of the professional life outside the academy.
The choice before us is thus not one of technology adoption but about whether we will design learning environments that support the capacity to think, to decide, and to act with discernment in a world where information is abundant, but understanding remains scarce. The answer will determine not only the sustainability of our institutions, but the capability of the individuals they prepare to lead, to innovate, and to serve.
Vistasp M. Karbhari is a Professor in the Departments of Civil Engineering, and Mechanical & Aerospace Engineering at the University of Texas at Arlington, where he served as President from 2013-2020. He is a Fellow and Board Member of Complete College America, and the co-chair of CCA’s AI Council,and can be followed on Twitter at @VistaspKarbhari and on LinkedIn at linkedin.com/in/vistaspmkarbhari.
- AI as a scaffold for learning: From access to judgment - May 11, 2026
- What higher ed can do about getting research into the K-12 classroom - May 8, 2026
- Building the AI-ready graduate - May 6, 2026