AI Is Firing Developers Everywhere… Should CS Students Worry?

Software Engineering in the AI Era: A Professor’s Note to His Students

TL;DR: Strengthen your fundamentals, learn how to work with AI, build small and ship quickly, communicate clearly, and do not wait for permission. A job is one path, but building on your own is another. You can start today.

1) Why this moment matters

If you are studying computer science in 2025, you are entering a very special moment. AI is no longer a side tool. It is becoming the default layer in almost every product and team.

It is easy to feel anxious when you read about layoffs in tech. That does not mean software engineers are no longer needed. The nature of the work is shifting. For those who know how to build, this is a season of opportunity, not decline.

2) The changing role of a software engineer

AI is very good at repetitive and structured work such as generating, refactoring, or testing code. So what role do we play as humans?

• Deciding what and why to build
  It is not just about writing code. You need to set direction, define success, and connect technology to real social and business value.
• Solving complex problems and thinking broadly
  You will often face ambiguity. The job is to weigh trade offs, combine ideas from different fields, and design solutions that hold up in the real world.
• Working with AI tools
  Think of coding assistants as teammates. Review their output, integrate it carefully, and adjust course when needed. Also consider how these tools can improve to better support builders like you.

3) Programming languages: where we came from and where we are now

The history of computing is also the history of programming languages. We moved from machine code to assembly to higher level languages that let us express more with less effort. Languages like Python and JavaScript are popular because they allow fast prototyping and productivity.

Low level languages did not disappear. C and C++ are still essential when performance and control matter. Programming GPUs with CUDA is one example. Many systems, compilers, and AI frameworks remain built with these languages.

AI coding tools are pushing programming closer to natural language. It looks like another stage in the evolution of programming languages. For now, traditional languages remain necessary because they provide precision and because they are the foundation that everything else stands on.

We cannot yet code entirely in natural language through AI. For the time being, you need to be able to work in both the traditional way and the new AI assisted way.

4) Core skills that endure

Trends come and go. Some skills do not lose their value.

• Computer science fundamentals
  Data structures, algorithms, programming languages, and systems basics. These allow you to understand and control AI generated code.
• AI literacy
  Know what LLMs, RAG, prompt engineering, and fine tuning are and when they make sense. Even if you do not train models, you should be able to integrate AI into reliable products.
• Domain knowledge
  Depth in areas like healthcare, finance, gaming, or robotics can turn a generic idea into a real solution.
• Communication skills
  Clear writing, clear speaking, and the ability to bring people together around a plan.
• Adaptability
  Learn quickly, connect new ideas to what you already know, and let go of approaches that no longer work.
• Quantitative reasoning
  Linear algebra, probability, and optimization build intuition for how models and systems behave.

A good practice question: What can you do that AI cannot?

5) How to study and practice

Do not obsess over tool checklists. Focus on rhythm and direction.

• Principle first learning
  Understand the underlying concept before grabbing a specific tool or library. If you learn the principle well, you can transfer that knowledge even when the tools change. For example, if you know how search algorithms work, you can adapt whether the code is in Python, Java, or C++.
• Balance self solve and AI support
  In the beginning, try to solve problems on your own. This builds intuition and judgment that no tool can replace. Later, once you are comfortable, bring AI into your workflow to test more ideas quickly or to automate repetitive tasks. Think of AI as an accelerator, not a replacement.
• Learn multiple languages
  Python is the foundation for much of today’s AI and data work, and it is also a flexible language for experimenting with ideas. C or C++ remain important when performance and low-level control are required. Each language gives you a different perspective that helps you understand the larger system.
• Build and share projects
  A portfolio is often just as powerful as grades. Internships also matter because they show you can work in real environments. Even small projects count if you finish them, write about what you learned, and put the code online. Employers and collaborators want to see evidence of growth, not perfection.
• Expect a stepwise growth curve
  Skill development rarely feels smooth. You will face plateaus where it seems you are not improving. That is normal. Use those times to review and deepen your understanding. Stay consistent and you will find that the next jump in ability comes suddenly, often when you least expect it.

6) Many possible paths

Getting a job is one path, but not the only one. You can become a solo founder or an indie hacker and ship to a global audience. This experience often makes you more attractive to employers.

Undergraduate research, open source contributions, community involvement, and side projects all open doors. Games, web services, and mobile apps are solid starting points.

Start small, ship quickly, learn from users, and improve as you go.

Conclusion

The AI era brings new challenges and large opportunity. What matters most is not speed alone. What matters is the ability to think deeply, adapt quickly, use AI wisely, and collaborate with others.

Programming languages are moving closer to natural language. Python, Java, JavaScript, and C/C++ remain essential. AI tools may feel like the new IDE. The vision, creativity, and problem solving still rest with you.

If you prepare along these lines, you will not only use AI. You will design and lead the future of intelligent systems.