The Absurdity of Eliminating Software Engineering from Computer Science: A Parody on Our Times

 The Absurdity of Eliminating Software Engineering from Computer Science: A Parody on Our Times

Srinivas.katharguppe

In the ever-evolving realm of computer science, a recent proposal has caused quite a stir among the academic elite. There is talk of eliminating the "mundane" subject of software engineering from undergraduate computer science curricula in favor of a greater emphasis on artificial intelligence and machine learning (AI/ML). This suggestion is not just an affront to the very pillars of computer science but also a veritable comedy of errors. Let us embark on a brief intellectual expedition into the heart of this quagmire.

In the immortal words of Frederick P. Brooks Jr., “The hardest single part of building a software system is deciding precisely what to build.” A sentiment that encapsulates the core essence of software engineering. It’s not merely about writing code; it’s about system architecture, requirements gathering, team coordination, testing, maintenance, and delivering actual value to end-users.

Dr. Grady Booch, another luminary in the field, famously said, "A fool with a tool is still a fool." We could fill our curricula with the latest AI/ML algorithms, but without the foundational knowledge of software engineering, our graduates risk becoming those very fools.

But let's humor this proposition for a moment. What if we did remove software engineering from the curriculum and focused solely on AI/ML?

The Integral Role of Software Engineering in AI/ML

The beauty of AI and ML is not just the algorithms or the computations, but how they are seamlessly integrated into scalable, robust, and efficient software solutions. Data preparation, feature extraction, model training, deployment, and maintenance are as much about software engineering as they are about algorithms.

As Andrew Ng, co-founder of Google Brain and Coursera, remarked, "Coming up with features is difficult, time-consuming, and requires expert knowledge. 'Applied machine learning' is basically feature engineering." This statement underscores the necessity of having strong software engineering principles to handle vast quantities of data, transform them, and structure applications around them.

The late Tony Hoare, a computer scientist who developed Quicksort and contributed to the development of ALGOL 60, once said, "There are two ways of constructing a software design: One way is to make it so simple that there are obviously no deficiencies, and the other way is to make it so complicated that there are no obvious deficiencies." This applies with heightened emphasis in AI/ML. Without the proper structuring and understanding of software engineering principles, our AI models can quickly become black boxes of inefficiencies and pitfalls.

Doomsday Scenarios: A World Without Software Engineering Education

Imagine a world where every AI/ML graduate crafts intricate neural networks but has no idea how to efficiently deploy them in real-world applications, leading to a digital realm rife with bloated, inefficient, and unreliable systems.

The healthcare industry, driven by AI diagnostics, could see patient records lost in the abyss of poorly structured databases, or worse, misdiagnoses due to inefficient algorithm deployment.

In finance, a poorly engineered algorithm could lead to catastrophic economic crashes, with AI trading systems making uncalculated decisions at breakneck speeds.

Let's not forget autonomous vehicles. Without rigorous software engineering principles, a minor bug in the system might translate into real-world accidents, putting countless lives at risk.

And as we lean more into smart cities and automation, a small oversight due to a lack of software engineering understanding could bring entire cities to a standstill. The chaos would be palpable.

In Conclusion

To even entertain the idea of sidelining software engineering in our computer science curricula is not just academically irresponsible, but it borders on the absurd. AI/ML, as brilliant and futuristic as it might seem, stands on the robust shoulders of software engineering principles.

As we move towards an AI-driven future, let's not forget the foundational knowledge that got us here. To do so would be to jeopardize the very progress we seek to make.

Disclaimer: The views expressed in this article are that of the authors and do not reflect the views of the magazine or any of the organizations that the author is associated with.