Have Computer Science majors lost their golden ticket to job security? As Artificial Intelligence (AI) enters the playing field in the realm of technology, talks of the unemployment crisis are at a peak. Subsequently, students are questioning the remaining value of a Computer Science (CS) degree and choosing to follow different paths. In contrast with Steve Job’s 2001 assertion that “everybody in this country should learn how to program a computer,” the 2025 reality is that any developer can go from idea, to code, to software all in natural language . So, the question arises: if any novice can prompt a chatbot to build a Flappy Bird clone in 30 seconds, then what are we doing taking CS courses like Fundamentals of Software Engineering and Theory of Computation? Since the rise of generative AI, the “barrier to entry” for coding has collapsed, yet the “barrier to excellence” prevails higher than ever. The truth of the matter is that the 21st century doesn’t mark the death of the CS major; it marks the evolution of a programmer’s role from syntax translator to systematic architect. Those who stay in the field are those who are ready to make the shift and break down the boundaries of human discovery by working alongside large language models.

Undoubtedly, AI has expedited the efficiency of countless industries, now providing just about anyone with the resources and tools to self-develop websites and automated systems. Raymond Fu , Data Specialist at Adobe, highlights that AI facilitates significant advances in possibilities within industries that rely heavily on software for their long-term goals. Large language models can now generate thousands of lines of code within seconds and translate seamlessly between programming and natural languages. Moreover, these models use high-level pattern-recognition skills to create user interfaces and automate repetitive tasks. A study by Github — the world’s largest software development platform — revealed that developers using Microsoft’s Copilot completed tasks 55% faster than those who didn’t. Notably, this productivity surge occurred even though the AI’s suggestions were not perfectly accurate. Furthermore, there was a statistically significant increase in the amount of people who successfully completed the assignment in the group working with Copilot versus the group without it.

However, while AI is one of the most powerful resources available for expediting the discovery process in the realm of CS, it doesn’t come without its limitations. One such drawback is its incapability of seeing the ‘bigger picture’ of a given project. As found by the International Conference on Software Analysis, Evolution and Reengineering (SANER) , when putting state-of-the art large language models such as Copilot Chat and Llama 3.1 to the test of fixing faulty code, most of its suggested solutions resulted in “compilation errors, test failures, or newly introduced maintainability issues.” When working with such tools, it’s important to note that their primary problem-solving skills come from predicting the next likely line of code, not fixing the current project. SANER’s findings emphasize that developer insight is crucial for creating usable, fully-functioning software.

This is where the role of professionals comes in. While AI generates the ‘bricks’ necessary to put together a project, the architects are the ones responsible for designing the blueprint and guiding the bricks. Architects understand the task’s purpose, the real-world context and scenarios in which a program may be implemented, and the long-term business goals and trade-offs. Most importantly, they know how to communicate and collaborate with others. Anyone can use AI to write code, but software engineers are still absolutely necessary when it comes to knowing how and where to implement it to get the most successful results. “We not only know how to prompt, but we also know what’s under the hood. The models, the data pipelines, the limitations, and risks,” says Fu on TedxCSTU . These understandings are critical, especially now that AI is being integrated into every product we’re using and building. Furthermore, “the next generation of AI is still built by software engineers.” By fine-tuning models, these scientists optimize performance and improve usability of AI to make it available and usable for everyone else.

As technological capabilities expand in the age of AI, software engineering isn’t just about syntax anymore; it’s about understanding logic. That’s why the shift in the industry redefines what it means to be a student in the lecture hall today. Software engineering is no longer a race to memorize, it’s a prioritization of systematic, deep thinking. While AI serves as a powerful assistant, future software engineers must step up to the challenge and orchestrate something great. It remains the responsibility of the human architects to refine these models and bridge the gap between prompt and product. For the students sitting in fundamental CS classes today, the challenge is clear: do not merely learn to use the tools, but master the logic that governs them. By understanding the fundamentals of the complex problems that AI cannot solve alone, you won’t just be surviving the era of machine learning — you’ll be the ones building what comes next. We are the ones who will remove the barriers and enter unexplored territory. We are the architects of the future.