A 30-Year Developer's Case Against GenAI in Software Work
A senior software developer with three decades of experience is rejecting the industry push to adopt generative AI tools for coding. The risk is too high, and the short-term productivity gains may create long-term dependency on tools controlled by companies with their own survival interests, not developers' best interests.
The developer, who has consistently reached senior levels at multiple employers, frames the decision as fundamentally about risk. GenAI tools like ChatGPT, Claude, and Copilot are being presented as necessary for career advancement. The counterargument: that adoption trades genuine skill for speed in ways that harm both individual developers and the profession.
Five Categories of Risk
Trust and addiction. GenAI chatbots are engineered to feel trustworthy through conversational design. The unpredictable nature of their output triggers dopamine responses similar to social media and gambling. Developers report surprise when answers are correct, anger when wrong, and fear when results exceed expectations. This cycle creates psychological dependency.
Loss of creative focus. Generative AI workflows interrupt the state of cognitive flow where expert developers produce their best work. Code flows from focused concentration the way language flows from writers. These new interfaces encourage multitasking, which the human brain cannot actually do. Interrupting deep focus trades innovation for convenience.
Only humans create truly novel solutions. Innovation requires deep expertise in a domain. Curating AI output instead of building that expertise prevents developers from knowing what constitutes good work.
Erosion of professional character. Developers lose patience for learning complex systems when prompted answers feel faster. Success in "agentic" workflows requires commanding chatbots like servants. Repeated practice of this dynamic-crafting harsh commands when tools "disobey"-shapes how developers think and behave toward others.
Eliminating debate with always-agreeable chatbots removes the friction that drives professional growth. Teams grow through disagreement and struggle. Mentorship-the transfer of tribal knowledge that makes teams competitive-becomes unnecessary when everyone relies on the same AI system.
Harm to junior developers. If experienced developers use GenAI and lose expertise, they cannot effectively mentor the next generation. Novices who use these tools skip the foundational learning required to become experts. The profession risks losing its pipeline of senior developers.
Corporate control. GenAI companies are not profitable at current pricing. They require massive adoption to justify their investments. Current subscription rates will rise significantly once dependency is established. The tools are neutral in appearance only; they are designed to increase corporate power over everyone who relies on them.
The Alternative
The developer plans to invest in human intelligence, creativity, and discernment. The outcome of GenAI remains uncertain. But the intentions of the companies promoting it are clear: they need widespread adoption for their own survival, not for developers' benefit.
If your organization is pushing GenAI adoption, these arguments merit consideration. The technology may deliver short-term speed at the cost of long-term capability. For development teams, that trade-off is worth examining carefully.
For those interested in understanding how AI tools fit into development careers, AI for Software Developers offers structured perspective on the topic.
Your membership also unlocks:
