Amid the AI Boom, Legacy Math Programs Face an Existential Reckoning
The global technology landscape is undergoing a seismic shift, driven by the relentless advance of artificial intelligence. What was once the domain of science fiction and academic research labs has erupted into the mainstream, becoming the core engine of innovation for industries ranging from healthcare and finance to manufacturing and entertainment. Governments around the world are scrambling to draft national AI strategies, pouring billions into research and development, acutely aware that leadership in this field is synonymous with future economic and geopolitical power. In this high-stakes race, the most critical asset is not silicon or capital, but human talent—individuals who can not only understand the complex mathematical underpinnings of AI but also wield them to solve real-world problems. It is within this crucible of technological transformation that a quiet revolution is brewing in university departments far from the gleaming tech campuses: the decades-old “Information and Computing Science” (ICS) program, a discipline built on pure mathematics, is being forced to confront its future.
For years, the ICS major has occupied a unique, if sometimes ambiguous, space in academia. Born in the mid-1990s at institutions like Peking University, it was conceived as a hybrid: a rigorous mathematical education fused with practical computing skills. The promise was compelling—a graduate who could bridge the theoretical world of equations and the applied world of software, capable of using scientific computing to model complex systems and develop specialized applications. The curriculum was, and often still is, dominated by advanced calculus, linear algebra, differential equations, and numerical analysis, supplemented by core computer science courses like data structures and algorithms. On paper, it produced versatile problem-solvers. In practice, as the digital world evolved at breakneck speed, many ICS programs found themselves adrift.
The problems are systemic and deeply entrenched. A primary issue is a profound identity crisis. What, exactly, is an ICS graduate supposed to do? The field’s breadth, spanning mathematics, economics, and computer science, became its Achilles’ heel. Many universities, in their eagerness to offer the program, failed to define a clear, market-relevant specialization. Students would spend four years immersed in abstract mathematical theory, only to graduate with a vague sense that they were “good with numbers and computers,” but without a concrete career path. This ambiguity led to a demoralizing lack of direction, where students moved from freshman year to graduation without a clear understanding of their professional value proposition. The result? A cohort of highly intelligent graduates who felt like “jack-of-all-trades, master of none,” struggling to compete in a job market that increasingly demands specific, demonstrable skills.
Compounding this identity crisis is a severe faculty gap. The typical ICS department is staffed predominantly by mathematicians—brilliant scholars with PhDs in pure or applied mathematics. Their expertise in proving theorems and deriving elegant solutions is unquestionable. However, their experience in the messy, iterative world of software engineering, data pipelines, and deploying machine learning models in production environments is often non-existent. Asking a professor whose research focuses on partial differential equations to teach a course on modern deep learning frameworks is like asking a classical pianist to produce a chart-topping electronic dance track. The translation is not impossible, but it requires a fundamental shift in skillset and mindset that many are neither trained for nor incentivized to undertake. Consequently, critical, cutting-edge computer science courses are often outsourced to adjunct faculty from other departments or, worse, simply not offered at all. This leaves a gaping hole in the curriculum, depriving students of the very tools they need to be competitive.
The course structure itself often reflects these underlying tensions. In a typical four-year program, the first two years are a gauntlet of pure mathematics—courses that, while foundational, can feel increasingly disconnected from the technological zeitgeist. By the time students reach their junior and senior years, when specialized, application-oriented courses should take center stage, the program often falters. Without dedicated faculty or a clear vision, the upper-level curriculum becomes a patchwork of electives with little coherence or progression. Core courses in areas like data mining, advanced statistical modeling, or AI ethics might be missing entirely. The consequence is a graduate who has a stronger theoretical math background than a Computer Science major but significantly weaker practical coding and systems-building skills. They are caught in an uncanny valley of employability, overqualified for entry-level data entry roles but underqualified for the high-paying AI engineering positions they aspired to.
Perhaps the most damning indictment of the traditional ICS model is its neglect of practical, hands-on experience. Theoretical knowledge, no matter how profound, is inert without the ability to apply it. Yet, many ICS programs treat laboratory work and industry internships as afterthoughts. The practical projects, if they exist, are often outdated academic exercises—building a simple calculator or simulating a basic physical system—rather than tackling real datasets or solving problems sourced from industry partners. The connection to the outside world is tenuous at best. Students graduate with beautifully typeset proofs and elegant algorithms sketched on paper, but they have never version-controlled their code on GitHub, never deployed a model to the cloud, and never presented their findings to a non-technical stakeholder. In today’s job market, these are not optional skills; they are the baseline requirements.
The rise of artificial intelligence has not created these problems; it has merely exposed and exacerbated them. AI is not magic. At its core, it is applied mathematics—specifically, statistics, linear algebra, and optimization—executed at scale by powerful computers. This is the ICS program’s natural territory. The irony is palpable: a discipline built on the very foundations of AI is being left behind by the AI revolution. The market is screaming for talent that can build and deploy intelligent systems, yet traditional ICS programs are still producing graduates optimized for a world that no longer exists.
This is the context in which a group of forward-thinking academics at Hebei North University, led by Zheng Wei, Mei Rui, and Li Zhenzhen, have proposed a radical overhaul. Their vision, detailed in a recent paper, is not to abandon the mathematical soul of the ICS program but to re-arm it for the battles of the 21st century. They argue for a complete reimagining of the program’s mission, curriculum, and pedagogy, transforming it from a theoretical backwater into a frontline producer of AI-ready talent.
The first pillar of their reform is a razor-sharp redefinition of the program’s goals. Instead of the vague “problem-solver” archetype, they propose a clear, market-driven mission: to produce graduates with “a solid mathematical foundation, strong innovation and entrepreneurship capabilities, and the proficient ability to apply mathematical models and computer technology for intelligent data analysis and processing.” This is not a subtle shift; it is a declaration of war on ambiguity. It tells students, faculty, and employers exactly what the program stands for and what its graduates can deliver.
To achieve this, the second pillar is a comprehensive curriculum overhaul. They advocate for a modular framework, but with a crucial twist: every module must be infused with AI-relevant content. This means introducing foundational courses on “Brain Science and Artificial Intelligence” in the early years to provide context and inspiration. It means replacing or supplementing traditional programming languages with Python, the lingua franca of data science and AI. It means making “Machine Learning” and “Deep Learning” core, required courses, not niche electives. And it means dedicating entire modules to “Big Data and Data Analysis,” ensuring students are fluent in the tools and techniques of the modern data economy.
But theory is not enough. The third pillar is a relentless focus on practical, experiential learning. The authors propose a structured, year-by-year “capability enhancement plan.” In the first year, alongside foundational math, students are introduced to the “why”—the real-world applications and career paths their studies can lead to. In the second year, they are pushed into “scientific research preliminary training” and “innovation and entrepreneurship activities,” encouraged to form teams and develop project ideas. By the third year, the focus shifts to intensive, hands-on “application ability training,” with dedicated “Python Training Weeks” and “Data Mining Training Weeks” that simulate real project sprints. Finally, in the senior year, the program culminates in mandatory, industry-aligned capstone projects and internships, ensuring every graduate has not just learned, but done.
The fourth and perhaps most challenging pillar is the deepening of industry-academia collaboration. The authors recognize that universities cannot solve this alone. They call for a new model of “win-win cooperation” with the private sector. This goes beyond the superficial “guest lecture” model. They envision a true partnership: bringing seasoned industry engineers into the classroom as co-instructors, jointly supervising student capstone projects that solve real business problems, and even co-developing market-ready software products. For this to work, trust must be built. Universities must move beyond viewing industry as a source of funding and start seeing them as co-creators of the curriculum. In return, companies gain access to a pipeline of uniquely trained talent—graduates who understand the “why” behind the algorithms, not just the “how” to call an API.
The implications of this reform are profound. If successfully implemented, it could breathe new life into a program that many had written off as obsolete. It positions the ICS major not as a competitor to Computer Science or dedicated AI programs, but as a crucial complement—a producer of “bilingual” technologists who can speak the language of both theory and practice. In an AI-driven world, where models are becoming increasingly complex and their societal impacts more significant, we don’t just need engineers who can build models; we need professionals who can interrogate them, understand their limitations, and ensure they are applied ethically and effectively. This is the unique value proposition of a mathematically grounded AI practitioner.
The road ahead is not without obstacles. Retraining faculty, redesigning decades-old curricula, and forging genuine industry partnerships require significant investment, institutional will, and a tolerance for risk. There will be resistance from traditionalists who see the infusion of AI as a dilution of mathematical purity. There will be logistical nightmares in coordinating with industry partners who operate on different timelines and priorities.
Yet, the cost of inaction is far greater. The AI revolution will not wait. If traditional ICS programs do not adapt, they risk becoming academic relics—museums of mathematical thought in a world that demands action. The work of Zheng Wei, Mei Rui, and Li Zhenzhen at Hebei North University offers a compelling blueprint for survival and, more importantly, for relevance. It is a call to arms for educators everywhere: to stop merely teaching mathematics for its own sake, and start teaching it as the powerful, world-changing tool that it is.
The future belongs to those who can harness the power of intelligent systems. The question for the ICS program is no longer if it will change, but how quickly and effectively it can transform itself from a bystander into a leader in the age of artificial intelligence. The time for introspection is over. The time for action is now.
By Zheng Wei, Mei Rui, Li Zhenzhen, College of Science, Hebei North University, Zhangjiakou, Hebei 0750002, China. Published in DIGITAL INSIDE Computer and Applications, 2021.10, Vol. 67-69. DOI: 10.3969/j.issn.1672-9129.2021.10.023.