Estimated reading time: 7 minutes
Key takeaways:
- Developers are facing a “psychological crisis” as AI masters technical tasks that previously defined their professional pride, craft, and social value.
- The industry is moving rapidly from a focus on “coding as a craft” to AI orchestration, leaving many struggling to internalize new roles that prioritize architecture over manual typing.
- Organizations risk alienating talent through rigid AI mandates and poor metrics. Success requires creating clear career roadmaps that treat AI as a collaborator rather than a replacement.
Developers are grappling with an identity crisis, as AI gets better at the thing they spent their lives mastering.
When software developer and researcher Annie Vella hit publish on a blog post titled The Software Engineering Identity Crisis, she expected it to only rack up a few hundred views, similar to earlier posts. Instead, it went viral, accumulating over 65,000 views and clearly striking a chord with developers around the world.
“I thought there’s going to be a lot of people out there feeling like the point of pride that they had in being the top person on the leaderboard on Hacker Rank, the best person to have written some recursive dynamic algorithm, well now that’s not really much to boast about when I could just command my Claude Code instance to do the same,” Vella said. “We’ve wrapped ourselves up in the quality of our work. Our ego and our identity are linked, for some so intrinsically that it’s got to have some psychological effects.”
Software developer Dave Gauer received a similar response to his essay titled A Programmer’s Loss Of A Social Identity.
“If I meet somebody out on the street and they’re like, ‘Hey, I’m a programmer,’ I’m like, ‘Whoa, that’s so exciting. I am, too. We’re gonna have so much to talk about,’” Gauer said. ”Now, that assumption does not hold, and I realized that’s actually a pretty big loss.”
Your inbox, upgraded.
Receive weekly engineering insights to level up your leadership approach.
An AI-driven identity crisis
For decades, identity in software development has been built around a combination of technical competence, problem-solving ability and the craft of writing code, said Carolyn Parry, a career identity coach. “We all tell stories about who we are, and it’s very firmly attached to identity,” she said.
An identity crisis – which is the disorientation of being asked to take on a new professional identity that hasn’t yet been internalized – can happen when it becomes difficult to tell that story with a sense of confidence, Parry said.
What makes this technological transformation particularly challenging is that AI is disrupting the industry at a faster pace than people have time to make decisions and make a switch, said Amy Chiang, a transformation coach and founder of Transforma Systems. Decisions, such as what degree to study or what to spend time specializing in, are often slow and deliberate, with certain assumptions about the future baked in. The current speed of change isn’t giving people that time to prepare, she added.
“Sudden shift is to put it lightly,” Gauer said. “We were arguing over curly brace style and tabs versus spaces, but more importantly, over the different paradigms: object-oriented programming versus functional, and all of a sudden, that’s out the window, and now we’re talking about, should we write code at all?”
Many developers got into the field because they want to write in the IDE, said software developer Simon Hoejberg, who penned the essay “The Programmer Identity Crisis.” Now developers are being asked to get out of the editor and do something very different, he said.
“One of the most disorienting parts about this is the quickness of the shift in culture,” Hoejberg said. “In the past 20 years, I’ve gone to programming conferences and, by and large, they’ve been about craft … so that for me has been the predominant culture in programming and that seems to have changed.”
More like this
Managing an identity crisis
Career coach Parry recommends that developers who picked roles specifically for the coding opportunities should now explore roles where they can feel anchored, regardless of how much it changes. But that’s easier said than done with developers contending with layoffs and a challenging hiring landscape as a result of cost-cutting and the emergence of AI.
Chiang, who’s been supporting transformation in organizations for decades, says it’s important to remember that resistance comes not just from a fear of a loss of a job, but also of financial stability.
Christian Chung, senior director of AI enablement at digital transformation agency Fueled, is observing the fears around skill decay and loss of craft in the developers he manages, particularly those in senior roles. As his company is positioning AI as a collaborator, he sees an opportunity to translate the love for the craft into solutions architecture.
“It matters less about how the code gets written,” Chung said. “It matters much more about how the code is architected, what problems it solves, how complete it is, how robust it is, how testable it is, how secure it is.”
“You can still affect the quality and outcomes without typing it out yourself,” he added.
Vella, who has worked both in individual contributor roles and in management, expects there to be opportunities for developers who don’t want to leave the code editor behind, but the work will be more about wrangling the AI using industry practices, patterns and coding standards, as well as tackling emerging challenges around performance, memory, and code auditing with AI tools.
Others may want to reclaim what was once the full scope of the role, which included tasks such as requirement gathering and architecture design, Vella said. The role of a software developer used to be much wider scope, but due to the demand for software and lack of supply, it became more specialized in recent years, she said.
Finding the right organizational culture
The creation of new roles or the broadening of existing roles will depend just as much on organizational culture as it does on the manager. Developers interviewed for this article said they were not opposed to the use of AI tools in development, but more the mandates and quantitative metrics surrounding their arrival.
“It’s incredible how much impact [vibe coding] is having on organizations,” said Hoejberg, who hasn’t personally been exposed to AI mandates but is observing them within the industry. “A CTO spends a weekend on something, and now a whole engineering org has to change because of that. It just seems wild for someone who’s not day-to-day coding to make that decision.”
Software developer Dan, who did not want to share his full name due to discussing sensitive information about a former employer, experienced this first-hand when he and colleagues had been using a simple code completion plugin, but were then forced to transition to AI-powered code completion, which was more powerful but far less dependable.
“It wasn’t just a change in tooling, but a change in doctrine,” said Dan, describing the company’s shift to an AI-first culture. “Switching from that autonomy to ‘no, you can’t use this thing.’”
Meta has implemented a new performance tool that connects ratings and incentives to AI-driven output, which, for developers, means tracking how many lines of code were generated with and without AI tools, as well as the number of errors or bugs associated with the developer’s code. But measuring lines of code has long been considered a poor way to measure productivity in software development. “It’s totally the wrong metric. Some of the best programmers delete more code than they write,” Gauer said.
Rather than mandate the use of tools or set metric goals, organizations should create more space to experiment with AI tools within guardrails, Chiang said. They should also be clear on what an employee’s career can look like with AI and how they will facilitate growth, whether it’s funding certifications or enabling them to try adjacent work in a low-risk way, she said.
“I really hate when you see leadership teams go, ‘Well, the AI is really good for you because it’ll free you up of all that tedious toil-y work that you were doing, and now you’ll be able to focus on higher-value work, or higher-level thinking,” Vella said. “Can anyone please explain what that work is?”
“Describe it so that they can imagine themselves doing it and can go away and start researching it and come up with some new ideas that they can bring back. Give them some hope, something to wrap their minds around,” she added.
Fueled’s Chung assuages concerns by creating individual roadmaps for developers with incremental goals on how they can integrate AI into their work. A roadmap can start with a simple task, such as conversing with AI, to then integration of AI into the code editor, to then branching out beyond the foundations to integrating MCP servers or building specialized agents.
It’s a way to make meaningful progress and garner wins along the way, Chung said.
An accelerated path to management?
As AI is viewed as a collaborator at Fueled, Chung uses his own transition from being an individual contributor to manager to help describe what the future of development work might look like. As a manager, he had to learn to derive value through the success of others and recognize his mentorship, prompts, and pushback all played a part in their success, which is much like managing an AI tool.
It’s also worth noting that direct reports tend to assume managers are coming from a place of stability, but that’s not the case in this environment, transformation coach Chiang said. They are navigating their own version of disruption with AI, while absorbing leadership pressure to adopt AI, she added.
“Under normal circumstances, middle managers rarely [have it] easy anyway and now this is just adding an additional layer of complexity,” she said. “But ultimately, naming it’s important, instead of pretending it’s not happening.”
London • June 2 & 3, 2026
New activities added to LDX3 London 🎉
