Mascot | Mascot | Getty Images
When Gemma Rogers started pursuing a computer science degree in 2002, software engineering was seen as a promising and lucrative career. This appealed to people who enjoy dabbling in the weeds of code, tracking bugs, and building programs from scratch.
Rogers, a UK-based software developer, told CNBC that the industry is always changing rapidly. But with the advent of artificial intelligence, the pace has accelerated, and AI tools are now part of the daily routine of every software engineer she knows. In a 2026 Stack Overflow survey, 83% of software developer respondents said they use at least one AI tool.
AI coding tools allow engineers to write, debug, and test code faster than ever before, but some are bemoaning the work of the past. Rogers has always loved the sense of accomplishment that comes from piecing together different pieces of code to find a solution. Especially if you’ve spent days or even weeks fixing stubborn bugs.
The sense of joy and enjoyment that I used to experience…it’s been taken away somewhat…it’s breadcrumbs of joy, not a full meal of joy.
gemma rogers
software developer
“The sense of joy and enjoyment that I used to feel when I had a problem and had to solve it has faded a little bit,” she says. “Now sometimes you can do it with a few prompts, which is great, but it’s a breadcrumb of joy, not a full meal of joy.”
Rodgers is not alone in feeling nostalgic for a time not so long ago. Software engineers say on social media that they miss the old way of working, where they could get into a flow state and write code for hours at a time.
Still, AI tools are useful and significantly lower the barrier to entry. “This industry is becoming an industry where a lot of people can succeed who wouldn’t necessarily have been successful 10 years ago, five years ago, even three years ago,” Rogers said.
It’s also important that the core skills engineers have learned on the job are not lost, she said. AI tools and internet services can go down unexpectedly, and engineers still need to know how to investigate problems and work on solutions themselves.
“Skills that we were taught years ago are being lost because (engineers) have to work so hard,” Rogers added.
AI code judgment
Work itself is also changing. James Broadhead, chief technology officer at AI decision-making company Taktile, told CNBC that engineers who used to take three weeks to build a prototype can now generate 10 prototypes in an afternoon. The most difficult part, he says, is “having the judgment to decide which is good.”
“The (AI) agent spits out so much code that (engineers) don’t read it all anymore,” Broadhead said. “Before, we could build up a sense of security and awareness about how the entire codebase was structured, but now the whole thing is constantly changing.”
AI skill gap
Joshua Vere, CEO of MindStone, a London-based AI training company, told CNBC that there is a growing gap between employees who know how to use AI effectively and those who don’t.
It’s a very, very real sadness… What you want to do is no longer something the company should be paying for at some level.
He said the industry used to talk about “10X engineers” who were dramatically more productive than their peers. Now, AI is widening the gap even further and producing “100X engineers.”
Wehrle believes that companies will increasingly have to decide whether to help their employees catch up with AI or prioritize those who can make the most of it. But I also understand why some engineers have a hard time adapting.
People who once excelled at this job often liked writing and working with code, and those people aren’t necessarily the top performers today, he said.
“It’s a very real sadness,” Wehle said. “What you want to do is no longer something the company should pay for at some level, because it’s no longer adding as much value to the company as it used to.”
machine supervision
However, not all engineers think that moving away from hands-on coding is a bad thing. Chris Sees, a software engineer with more than 20 years and CEO of London-based virtual office company Hoxton Mix, told CNBC that AI gives developers more time to work on interesting problems. He sees AI as a talented young engineer who still needs supervision.
“The key is knowing what to delegate and when human judgment is needed,” Cease says. “It would be foolish to think that AI will get everything right. We know there is great satisfaction in proving and identifying where AI is wrong.”
Still, Sees worries about what will happen to the next generation of engineers if AI takes over the day-to-day coding and debugging that they have traditionally learned. These tasks are “where lessons are learned and scar tissue is built,” he said, and companies need to ensure that AI becomes part of the learning process rather than a shortcut.
For Rogers, it’s important to be able to control how technology is used. As a self-employed developer, she can use AI to automate the boring parts of her projects while freeing herself to solve more interesting problems.
She believes there is something important about solving difficult problems rather than immediately outsourcing them to technology. Friction and discomfort are an important part of the human experience, she said.
“As humans, we need to feel like we’ve accomplished something, and accomplishment requires a challenge,” she said. “If you can’t practice that in many aspects of your life, you won’t be able to bring out the best in other people’s good feelings.”
