Developer warns AI reliance is eroding core coding proficiency and professional standards

A software developer has reported that heavy reliance on artificial intelligence for writing and coding tasks is significantly diminishing their personal technical abilities. The author, who has utilised AI tools for approximately two years, states they have effectively forgotten how to code manually and are currently in the process of relearning the skill by hand. This admission highlights a growing concern within the technology sector regarding the cognitive impact of delegating core creative and logical tasks to generative models.

The developer described the temptation to use AI for articles, code, and documents as "god damn tempting," noting that their previous self-assessment as a "somewhat talented... well... mediocre software developer" has been overshadowed by a perceived decline in their own capabilities. After spending the past one to two years entirely prompting AI systems and not writing a single line of code manually, the author expressed that the loss of coding proficiency is "sad and depressing," given that coding "used to be my life."

Beyond technical atrophy, the author noted that AI-generated text lacks their personal voice and exacerbates feelings of imposter syndrome. The developer observed that AI output often fails to convey the intended message accurately, reading as distinctly artificial rather than authentic. This reliance on external validation has led to a cycle of self-doubt, with the author catching themselves about to submit their own writing to Claude to check for sense or flow, driven by a fear that their own work is insufficient.

Despite these personal challenges, the developer suggests that the profession of software development will not disappear. They argue that while AI may reduce the overall demand for developers, there will still be a necessity for individuals capable of reading and writing code. This shift could potentially reverse a long-term trend observed over the past 20 to 30 years, where high demand for software developers led to a dilution of professional standards as the barrier to entry lowered.

Referencing software engineering figure Robert Martin, also known as Uncle Bob, the author contextualised this potential shift within the history of the field. Martin has previously lectured that before computer science became a mass-market profession, programming was dominated by physicists, mathematicians, and academics. The developer hopes that the current technological landscape might restore a higher level of professionalism to the industry, countering the decline in standards that occurred as the demand for developers skyrocketed.