Estimated reading time: 8 minutes
Key takeaways:
- A viral “I’m done with AI” video hit 900k+ views, striking a nerve with devs feeling the same grief.
- Reasons cited include eroded craft, ethical concerns, and real mental health costs.
- Some are already paying for it: lost clients, missed jobs, and one left the industry entirely.
Developer Brett Chalupa didn’t tell his boss or coworkers he’ll be abstaining from using AI going forward. He told the whole internet in a YouTube video.
“My career, my craft, my hobby has been drastically changed by the introduction of AI-coding tools,” he began the now-viral video before declaring he’s finished coding with AI (and using AI altogether).
Chalupa, who’s been coding for 20 years, went on to explain how the shift to AI coding has left him feeling disillusioned and purposeless, why he thinks it’s bad for software and business, and his concerns about AI’s larger impacts on society.
“Whatever comes of this decision – if I get fired, if I can’t find a job, if my career as I know it is over – so be it,” he concluded.
Chalupa merely said out loud what many developers are feeling. While some have been enthusiastic to embrace AI, others are grieving the work they loved and questioning their place in the industry.
At the time of writing, Chalupa’s video has garnered over 900,000 views and 10,000 comments, most of which relate to his feelings, with some even thanking him for having the courage to speak up. Chalupa says he’s also been inundated with supportive messages.
Your inbox, upgraded.
Receive weekly engineering insights to level up your leadership approach.
Of course, not everyone who feels this way is able to risk their livelihood by taking such a stance, but Chalupa is not alone. Other developers are similarly deciding they won’t adopt AI, even if it costs them their career. If the craft is removed, they’re not interested anymore, and they don’t care if they get left behind.
“I’ve already effectively left the industry,” says Anthony Manning-Franklin, who recently departed his CTO position after the company’s founder insisted he did not want any engineers coding by hand. “I’m not interested in coding with AI, and I don’t want to work with anyone that codes with AI.”
Why these devs are rejecting AI
Manning-Franklin gave AI coding a fair chance, even building his own neural networks from scratch without any libraries so he could understand it from the bottom up. He says he’s found AI to be more of hindrance than a help, putting out massive quantities of code that needs to be refactored or reworked.
“I have, far too many times, refactored a 1,000-line file down to almost 100 lines, or a 3,000-line solution down to 250 lines. Digging through 3,000 lines of damp spaghetti during an incident is depressing enough, but it’s much worse when it’s brand new code that no one wrote and no one actually understands,” he says.
Manning-Franklin also feels AI erodes developers’ skills and finds it impossible to reconcile using it with his values regarding the environment. The developers choosing not to use AI that LeadDev spoke with all cited these reasons, along with personal job dissatisfaction and concerns about AI’s larger negative impacts on society.
Pieter Groenendijk, a developer in the Netherlands who just completed his computer engineering degree this year, is also refusing AI. He notes that 40% of the Wikipedia page for AI is dedicated to ethical issues.
What’s more, AI coding lacks the sense of craft that sparked his interest in the first place. At his core, he likes writing the code, and he doesn’t want to outsource his thinking. Even when Github Copilot first came out and performed like smarter autocomplete, he didn’t like waiting for the AI to finish his thoughts.
“Of course, it’s really different now,” he says. “For me, I would say it’s more of the things I don’t like.”
Freelance developer Jared White says he regularly researches and keeps up with the latest AI tools, but “can’t get over [his] ethical objections to personally try using them.” These objections include environmental damage from data centers, AI eroding trust in society, and the fact that AI models were trained on scraped data without consent.
He’s discovered tons of the open-source code he’s contributed over the years in training datasets, he says, and doesn’t think that should be allowed.
After Chalupa’s employer mandated AI adoption, he used it to code in earnest for around 18 months. He even liked it initially and felt like it was the future.
Then, he became increasingly alarmed by the societal impacts – environmental damage, chatbot sycophancy, etc. – but primarily, he refuses to use AI because of the effect it’s had on his own mental health. He describes feeling depressed and struggling with existential dread, unsure of his purpose and unfulfilled by the idea of prompting AI models.
“I realized through this 18-month journey that this is making me feel horrible. The mental health cost is way too much for me,” he says.
The results of a study on the psychological costs of AI adoption in software engineering, published on arXiv, reflect these feelings. In analyzing a large software development services company a year after it launched its AI adoption, the researchers found widespread “craft identity disruption, meaning and satisfaction erosion, cognitive and workload intensification, and uncertainty distress.”
Chalupa acknowledges shipping more code when using AI, but he thinks the efficiency tradeoff is a mistake. He’s found AI causes developers to lose their grip on how the codebase is shifting and changing over time. He also feels apathetic toward code the AI wrote, which he says is bad for business.
“The writing of the code – that process where you’re in the zone, you’re thinking, ideas are coming up, you’re making these micro decisions, and that rhythm – I think that’s a beautiful, creative, enjoyable process,” he says. “Just reviewing what the large language model (LLM) spits out is the opposite of that for me.”
More like this
Loss and grief
Chalupa works for a small company where token usage isn’t monitored. Quitting AI hasn’t been an issue for him at work so far (he knows colleagues have seen the video), but realistically, he knows this could backfire on his career. The biggest risk as he sees it is that he gets fired and can’t find another job in the field, since he imagines anyone hiring would expect him to use AI.
A year ago, White was already starting to feel the impact of his strict “no-genAI policy,” including parting ways with a long-time client after they went all-in on agentic coding. More recently, he missed out on a full-time role that otherwise felt like a great fit after it became clear in the interview that the company was “San Francisco-startup-style-full-steam-ahead” on AI coding. He and the interviewer were both disappointed by the outcome, he says, as they clicked personally and he has the specific skills they wanted.
“It’s always frustrating when you feel like you can bring value to a project and get behind the mission, and then to be told, ‘we have a particular way of structuring the tools that get used by software developers here, and if you don’t fit into that particular structure, you’re not welcome.’ I feel like I’ve had so little of that at any other time in my career,” he says, adding that this “dictatorial approach” to how developers structure their work feels like a massive cultural shift.
White has been looking into other career options and starting to make peace with the fact that he may not be able to maintain an income as a software engineer. He’s also been building a community for developers who feel similarly about AI.
He rebranded the developer Discord he’s run for years into the “Human Web Collective,” geared toward developers who don’t like using AI tools and any other creatives who feel squeezed out by the technology. He also launched handcraftedcode.org, a directory where developers who still code the old-fashioned way can add their information (and potentially find work opportunities with like-minded people). He’s under no illusion they’re going to spark such a revolution that everyone will be back to coding by hand, but he wants to prove that “this bifurcation is real and that it means something.”
Berlin • November 9 & 10, 2026
Close the gap between what leadership expects and what’s actually possible at LeadDev Berlin.
Overall, these developers all say they feel grief. “Grieving my career and a craft that I loved has been significant and painful. I don’t even enjoy contributing to open source anymore, because I feel like any of the tools I create aren’t going to be used by a real person, if they’re even discovered at all,” says Manning-Franklin.
Chalupa says he believed he had a career path he’d be happy in for the next 30 years. Going from that to wondering if he wasted so much of his life learning a skill that “a stupid chatbot can do” has been massively destabilizing.
For Pieter, that sense of loss is not for the career he loves and is losing, but rather the one he never got to experience. Since he’s just starting out, he says it feels like he isn’t getting the chance to use what he spent years learning.
Even so, he sees abstaining from AI as the best choice because he wouldn’t be happy working as a programmer if he used AI. He says he’s aware he’s coming from a position of privilege and can take the stance because he has the safety net; he’s not at risk of becoming homeless, and his country provides universal healthcare. He’s currently working on his own software project, and if it comes to it, will look for another career.
“I have been grieving a little bit. Sometimes I’ve just been very down thinking, what am I gonna do?” he says. “Eventually I overcame that, mostly because I decided I should stop using it. When I made the choice, that’s when I kind of started getting my soul back.”
