AI will not immediately replace computer scientists. Here are 10 reasons

As AI systems expand their already impressive capabilities, there is an increasingly popular belief that the field of computer science (CS) will soon become a thing of the past. This is passed on to future students today in the form of well-intentioned advice, many of which, despite their intelligence, are mere hearsay from individuals who speak outside their expertise.

Famous figures like Nobel Prize-winning economist Christopher Pisarides have made this argument, and as a result, he has taken root on a much more mediocre level. I have personally heard high school career advisors dismiss the idea of ​​studying CS completely despite having no knowledge of the field itself.

These claims usually share two general flaws. The first of them is that advice comes from people who are not computer scientists. Secondly, there is widespread misconception about what computer science is actually involved.

Myth of AI and code replacement

It’s not wrong to say that AI can create computer code from the prompts so that AI can generate poems, recipes and cover letters. It can increase productivity and speed up workflows, but none of these eliminate the value of human input.

Writing code is not synonymous with CS. You can learn to write code without attending a single university class, but your CS degree goes far beyond this one skill. Among many others, it includes complex engineering systems, designing infrastructure and future programming languages, and verifying systems for ensuring cybersecurity and accuracy.

AI will not be able to ensure these tasks, nor will it be possible in the near future. Human input remains essential, but pessimistic misinformation risks piloting tens of thousands of talented students from important and meaningful careers in this critical field.

What AI can and cannot do

AI is good at making predictions. Generated AI enhances this by adding a user-friendly presentation layer to your internet content. This rewrites the information, makes the abstracts and formats resemble human works.

However, today’s AI is not truly “thinking.” Instead, it relies on logical shortcuts known as heuristics. This means that despite speaking like a person, it cannot reason, feel, care or desire anything. It does not function in the same way as the human mind.

Until recently, it seemed that “quick engineering” would replace CS. However, today there are few rapid engineering jobs, but companies like LinkedIn report that CS expert responsibilities have actually increased.

There’s a lack of AI

What AI offers is a more powerful tool for CS professionals to do their job. This means that you can move the concept further from ideas to market development when there is less support roles and technical leadership is needed.

However, there are many areas where professional human input is still essential, whether it is the need for trust, surveillance or human creativity. There are many examples, but there are ten areas that stand out in particular.

Adapt hedge fund algorithms to new economic situations. This requires a deep understanding of not only code but algorithm design and market. Diagnosis of intermittent cloud service outages from providers such as Google and Microsoft. AI can troubleshoot on a small scale, but it cannot contextualize high stakes troubleshooting at a large scale. Rewrite the code for a quantum computer. AI can’t do this without extensive examples of successful implementations (it doesn’t exist now). Design and protect new cloud operating systems. This includes high-level system architectures and rigorous testing that AI cannot perform. Creating energy-efficient AI systems. AI cannot spontaneously invent low-power GPU code or reinvent its own architecture. Building safe, hacker-proof, real-time control software for nuclear power plants. This requires mixing the expertise of the embedded system with code and system design translation. Ensure that the surgical robot software works under unpredictable conditions. Highly secure verification exceeds the current scope of AI. Design a system to authenticate email sources and ensure integrity. This is a cryptographic and interdisciplinary challenge. Auditing and improving AI-driven cancer prediction tools. This requires human monitoring and continuous system verification. Building next-generation safe and controllable AI. It is not possible for AI to evolve towards safer AI by itself. This is human responsibility.

Why Computer Science is Still Essential

One thing is for sure, AI will reconstruct how engineering and computer science are done. But what we are facing is not the wholesale destruction of the field, but the change in the way we work.

Every time you face a whole new problem or complexity, AI alone is not enough for one simple reason. It depends entirely on historical data. Therefore, maintaining AI, building new platforms, developing fields like trustworthy AI and AI governance all require CS.

The only scenario that may not require CS is when you reach a point where you no longer expect a new language, system, tool, or future challenges. This is so unlikely that it will never disappear.

Some people argue that AI could ultimately perform all these tasks. It’s not impossible, but even if AI is at this advanced level, almost all occupations are at equal risk. One of the few exceptions is those who build, control and advance AI.

There is a historical precedent for this. During the Industrial Revolution, factory workers were expelled at a 50-1 ratio as a result of rapid advances in machinery and technology. In that case, the workforce actually grew in a new economy, but most of the new workers were those who could operate or modify machines, develop new machines, focus on machines, and design new factories and processes.

During this massive upheaval period of technical skills, technical skills were actually most in demand, but at least not. Today, parallelism is true. CS technical expertise in particular is more valuable than ever before.

Don’t confuse the next generation with the opposite message.

This article will be republished from the conversation under a Creative Commons license. Please read the original article.