Rethinking Support - How HiTA Makes Constructivist Learning Scalable
Since 2001, I've taught STEM courses at the university level—first computer programming, then applied and computational mathematics. During this time, I've witnessed two significant technological shifts that fundamentally altered how students access and interact with information. First was the digitization of content originally found in dense texts, decoded by experts in post-secondary education environments, which is now instantly accessible through the internet and further consolidated into Wikipedia. The second shift is currently underway, driven by the emergence of knowledge-based generative AI systems, which provide flexible on-demand content access and adaptable, less passive student engagement. Through both transitions, the core question that educators wrestle with remains unchanged: What does education and learning look like when consumers have near-limitless access to academic content?
Universal Design for Learning (UDL) offers a powerful answer to the persistent challenge of educational equity. Rather than treating student differences as obstacles, UDL recognizes variability as the norm and proactively designs learning environments to accommodate it. At its core, UDL is a research-based framework that advocates for flexible approaches to presenting information, engaging students, and assessing understanding, thereby ensuring that all learners, regardless of their background or ability, have meaningful opportunities to succeed. UDL posits that when learners are offered multiple means of engaging with material, various ways to access and comprehend content, and diverse options for expressing what they know, all students are empowered to progress through the stages of cognitive development described by Bloom’s taxonomy. For the first time in my career, due to the help of HiTA, this ideal seems truly possible.
Though UDL addresses the subjective needs of the naturally varied distribution of learners encountered in education, self-regulated learning (SRL) theory describes the process by which students actively set goals, choose strategies, monitor their progress, and reflect on their learning to take greater control over their academic success. Here, there are three critical phases that students must navigate: the prospecting and planning phase, where they set goals and develop learning strategies; the performance phase, where they monitor their progress and adjust their approaches; and the reflection phase, where they evaluate their learning and plan next steps. For educators, it's not hard to see the need for learning environments that allow for both universal design and support through regulatory cycles, but the primary roadblock has been capacity.
As the internet subsumed content, my educational approach shifted to emphasize the use of critical reasoning skills to break down complex contextualized problems into standard textbook procedures/algorithms. Students entering the industry, now complicated by the complexities of rapidly growing technologies, were served well by this approach. However, for some students, the difficulty of quickly synthesizing content knowledge to decode dense problem statements was too great, and I was left with the problem of supporting those students who were investing significant time and cognitive energy, and mainly grouped into two cohorts defined by those who wanted more complexity and those who wanted less.
While these groups may seem fundamentally opposed, their underlying goals often overlap in surprising ways. Both are ultimately seeking solutions that better fit their learning preferences and regulatory workflows, even if their approaches differ. Knowing that there is no one-size-fits-all solution to this issue, implementing the UDL principles, providing multiple means of engagement and expression, seemed like a solution. However, the operational demands of orchestrating complex, individualized learning experiences seemed insurmountable. This is precisely where educational AI has proven transformative, not by changing pedagogical goals, but by making them achievable at scale.
HiTA has enhanced my capability to support the learning process in ways that were previously impractical. For example, in our general mathematics education courses with standard assessment structures, e.g., written homework and timed examinations, I've found HiTA helpful for administering frequent low-stakes activities that help students prepare for class content and practice rudimentary, but necessary, calculations. Additionally, due to the flexibility of the HiTA system, I've mirrored this process with activities that prompt students to consider their actions and identify areas for growth. Lastly, with HiTA's ability to incorporate course-specific materials in its responses, e.g., worksheet files, practice examinations, slideshows, pictures of lecture boards, transcriptions of lectures (private), students have immediate, relevant support at their fingertips while engaging in course activities. With HiTA’s guidance available at every step, students are empowered and encouraged to take greater ownership of their learning process, fostering self-regulated learning habits that are essential for both successful course outcomes and lifelong learning.
Outside of general education coursework, where curricular elements and assessment techniques are less regimented, HiTA's support has been indispensable for advising student-created mathematical modeling projects, assisting me in providing feasibility checks, level-appropriate background information, broader context, and computational strategies to student groups. Beyond this, it directly assists student groups in brainstorming, task management, and the development of large-scale codebases in computational environments such as Python, MATLAB, R, and Mathematica. HiTA's support makes me feel comfortable guiding student projects in topics well outside my areas of expertise, allowing student groups to have autonomy and choice when choosing a modeling topic, a fundamental aspect of UDL. This technological shift has led to remarkably sophisticated student projects, e.g., models for identifying AI-generated voices, exploration of murmuration in flocks of birds, development of gameplay strategies for challenging human puzzles, agent-based desired path modeling, and prediction of course outcomes using data on student achievements and behaviors. With HiTA, we can leverage a vast amount of knowledge coupled with code building, allowing for authentic applications of mathematical and statistical concepts such as dynamical systems, Fourier analysis, data analysis, statistical modeling, and Markov chains, in contexts that feel relevant and exciting to the learner.
Regardless of technological advancement, the response from educational institutions remains fundamentally the same: support the development of critical reasoning and problem-solving skills in our learners. What I've discovered through implementing AI-enhanced educational approaches is that technology doesn't change what good education looks like, it makes excellent education more achievable. The pedagogical principles I've always believed in, like UDL and self-regulated learning support, remain as relevant as ever. What has changed is my capacity to implement these principles effectively with larger numbers of students across more diverse learning contexts, as HiTA systematically addresses the barriers that once prevented me from fully realizing UDL/SRL in my teaching, e.g., flexibility in the access/representation of created content, the need for expertise across multiple domains, and the difficulty of providing individualized feedback before and after high-stakes assessment, making it possible to deliver individuated educational engagements meeting students at their point of need. With tools like HiTA, the promise of educational equity and excellence is within reach.
About The Author
Scott Strong