How HEIs can keep pace with a technology that constantly reinvents itself

Recently Google turned its NotebookLM into a full AI workspace and launched Gemini 3 Flash as the new default model. The open-source world dropped GLM 4.7, which is currently crushing every coding benchmark ever. ChatGPT released GPT 5.2, creating benchmarks in the long content format.

Constant change

Artificial Intelligence is no longer an emerging subject. New models, frameworks, tools, and ethical debates surface every day. Yet, most universities operate within a syllabus-driven system where curriculum revisions take years, not months. According to the FICCI–EY–P AI Adoption Survey 2025, 86% of students in Indian higher education institutions already use AI tools in their studies but only about 57% of those institutions have formal AI policies in place.

So, how do you teach a technology that reinvents itself every 12-18 months within an academic structure designed for stability? The answer lies in a fundamental rethinking of how AI is taught, assessed, and continuously refreshed within the university ecosystem.

First, learning needs to shift from being tool centric to principle centric. To remain competitive, universities need to focus on teaching the general principles of computing. While there may be different tools, the core principle upon which they were developed, how they will work and the concepts behind them, must remain consistent over time. Institutions should, therefore, emphasise the development of algorithmic thinking and decomposing problems; an understanding of both data literacy and statistical reasoning; interpretation of models, their limitations, and their ethical implications; and the ability to consider a systemic perspective and how students interact with AI. By emphasising these basic tenets, institutions can prepare their students to individually adapt as new technologies are created.

The second involves modular and stackable curriculum design. AI education should not be approached as a single block of learning. Institutions may revise short modules, electives and micro-credentials every year to include developments in the AI field, while maintaining their existing curriculum’s core structure. This also encourages cross-disciplinary collaboration by incorporating AI components into business, healthcare, humanities or design programs while remaining consistent with how AI operates within these fields.

A case study from Zhejiang University in China adopted a dual-module framework (20-hour AI training + embedded ethics discussions), introduced deep learning models, LLMs, AIGC, LoRA, and ComfyUI while maintaining the original curriculum structure, supported by dedicated technical instructors. It is an example of how phased, modular approaches can enhance student skills without overhauling existing degrees.

Faculty development

Third is faculty enablement, an aspect as critical as student curriculum. A quickly evolving field requires ongoing support for faculty development. Universities need to actively support their professors through faculty development programmes that focus on the practical applications of AI, through collaborative teaching models that bring together both practitioners and educators, and through research sabbaticals that allow faculty to work at applied research labs or within industry settings. If educators are up to date with technological advancements, students will also benefit by receiving an education that incorporates these advancements.

Finally, universities must see themselves as learning ecosystems, not content providers. In our rapidly changing world, AI technology is advancing so quickly that the half-life of knowledge is decreasing rapidly, and the education process should not end at graduation.

In addition to providing new exciting programmes for graduates, universities should also offer an opportunity to continue learning by providing access to updated modules (the same as a continuing professional education), refresher courses, and emerging research. The institutions that fully embrace this vision will not just be able to keep pace with the rapidly changing technological world, but will also be able to shape it.

The writer is the founder and CEO of Futurense.