As the saying goes, a long time ago God declared: Blessed are the meek for they shall inherit the earth. In the 21st century, the word Meek was replaced by the word Geek. It rhymes well. This reflects the reality. The geeks have actually inherited the Earth because they control the layer that controls us. They manage software, smart machines, and data centers. And this gives them strength. Like the esoteric Egyptian priests of old, they speak a language only they understand. This is the language of software programming. While normal people see only abstract symbols, letters and numbers on the screen, software developers see logic and functionality. Now, they have a contest called Cloud Code.

Cloud code has been the talk of the social media town in the last 10 days. In tech circles, this new AI-powered coding tool has aroused various emotions. Some people are surprised by its coding capabilities. Some people are scared to see how good AI has become. Some people are surprised and worried about the uncertain future of software engineering jobs. Almost everyone agrees that cloud code is the real deal, that AI is coming for tech jobs, and that the world of software engineering is undergoing a major transformation.

An

AI tools like Cloud, ChatGPT, Gemini are so good that they will do the work of many knowledge workers. Recently Elon Musk said that AI is already so good that it can replace 50 percent of human workers in white-collar jobs. But out of all the fields, coding and software engineering is one area where it seems AI will prove to be a bigger challenge for human workers. As the joke goes, it is somewhat ironic that the software engineers working on AI are actually creating a monster that will take their place.

Cloud code is very good at logic and hence coding

It didn’t look like this a few years ago. At the time when ChatGPT came on the scene – with ChatGPT 3 – it seemed that the first jobs that AI would take over from humans would be ones where writing was the main task. Three years later, it has kind of impacted jobs. AI can actually write reports and memos decently. It can also perform quite well on design-related tasks. Yet, one area where it is proving to be most disruptive is software engineering.

This may seem like a surprise but it is not. Software engineers realized this a little late, perhaps because their god-like abilities blinded them, but coding for AI is one of the easiest tasks because it lacks the nuances that many other jobs require. Instead, coding is all about logic and functionality. Just mere efficiency. Most humans are not good at it because they are not good at logic and mathematics. This gave software engineers, who are good at remembering procedures and syntax along with logic, a huge advantage over the rest of the world. But what made some humans good at software engineering is exactly the kind of ability that AI also excels at.

AI is (still) bad in terms of nuance, aesthetics, and abstract ideas. It still lacks problem solving capabilities. And by problem solving I don’t mean solving: using Euclid’s algorithm, find the HCF of 2048 and 960. Instead, AI cannot yet solve a problem that involves the human factor. For example, it cannot plan (feasible and good) to build a road that will be used by 10000 humans every day with different types of vehicles in different weather conditions.

But software engineering? Well, that’s where AI makes it better, as Cloud Code is showing this week. There are two reasons that explain why software engineering has increasingly become a field where AI is ready to take over from humans.

Coding, when expressed in languages ​​like C++ and Python, may seem daunting and scary to most humans, but really it is just logic flowing step-by-step. And often, there is already a predetermined path from one step to the next, at least for junior and mid-level software engineers. This means that most software engineers – but not all – simply have to write this path in coding language so that it can be understood by machines. Simple logic based on mathematics, which is either true or false, makes it easy for AI to replicate the coding. There is no need to deal with specifics here. The code either works, or it doesn’t. If it works then all is good. If it isn’t, it isn’t.

This is different from things like medicine. A doctor has to go into details. There is almost always nothing binary about medical treatment. There are advances, there are loopholes, there are compromises, there are nuances. AI can help doctors, but not replace them.

Coding doesn’t even require human touch

Second is the lack of aesthetics and human touch. Or rather, their absence does not matter to a great extent. When you write a report, let’s say a report on market conditions for selling cotton in Honolulu, your report can be unique in a way that anyone who reads it will understand. It’s not the same with coding. Although it may be unique, it does not matter to its users. The aesthetics of the code is not important, only its functionality is important. Whether you write this code elegantly or elegantly, it will not be of concern to its users. However, the exception to this would be UX and front-end. But even there, by its nature, software is more important because of its functionality rather than its looks.

AI can force its way through coding to provide functionality and it won’t make a difference. By its nature, software engineering does not value the human touch. After all, he doesn’t need it. With AI now coming onto the scene, human software engineers can no longer use their greatest strength – the human touch – to protect their jobs.

So, what comes next? Is it really over for software engineers? There is definitely going to be disruption. But how this all plays out depends on a number of factors, including policy factors. Tools like cloud code can make software engineers more productive, more diverse because they allow them to use and benefit from AI in ways regular people can’t, just as regular people can never code. New technological jobs may open up in the future, jobs that require human reasoning and abilities. For example, hunting for bugs in code created by AI, a skill that requires abstract methods and the ability to think with nuance. Or jobs that require systems-level thinking with all their complexities before the hard work of coding can be put on AI. Or AI tools could actually take jobs away from humans.

Of course, all this is going to happen in every sector that needs white-collar workers. All knowledge and creative work is going to be influenced and influenced by people like ChatGPTT, Gemini and Cloud. But in the near future – say in 3 to 5 years – the main area that will see a clash between humans and AI will be software engineering.