3 hrs ago
Engineer Says Years of AI Use Weakened Their Coding Skills
A software engineer said they now use AI for much of their coding work.
They started by asking it to help with small problems, but gradually relied on it for bigger tasks too.
Now, they sometimes have trouble explaining how their own code works.
They said this makes them worried about their skills and their job.
Some people online said busy deadlines can make it hard to learn every detail.
Others advised checking AI-written code and doing some work without AI.
One commenter said the most important thing is making products people find useful.
Another warned that code people do not understand can lead to bugs.
A software engineer with three years of experience said they had become heavily reliant on AI tools for coding.
The engineer said they now struggle to understand or explain code they submit, especially when senior colleagues ask about it.
They described diminished focus and confidence, and said they find it harder to read documentation or trace code manually.
Some commenters said deadline pressure can leave developers little time to understand their work; others recommended scaling back AI use and verifying generated code.
Other commenters argued that delivering useful products matters most, while one warned that overreliance can contribute to bugs and regressions.
- Who
- A software engineer with three years of experience, along with people who commented on the engineer's post.
- What
- The engineer described becoming dependent on AI for coding and struggling to understand and explain generated code.
- Where
- The engineer described the issue in a social media post; the platform was not specified.
- When
- The engineer said this developed over three years; the post's exact date was not stated.
- Why
- The engineer attributed the problem to relying on AI for much of their coding work as tasks grew more complex.
Concerns About AI Dependence
Practical Benefits and Uses of AI
Effects on developers' understanding
Concerns About AI Dependence
The engineer and one commenter warned that relying heavily on AI can weaken understanding and contribute to bugs or regressions.
Practical Benefits and Uses of AI
Another commenter said AI can help developers deliver features and products, arguing that the value delivered to customers is the key measure.
How to use AI at work
Concerns About AI Dependence
The engineer's experience illustrates concern that accepting AI-generated code without fully understanding it can leave developers unable to explain their work.
Practical Benefits and Uses of AI
A commenter described a more limited approach: letting AI write code and tests, then verifying the code and manually testing and debugging it.
Role of workplace deadlines
Concerns About AI Dependence
Commenters warned that pressure to meet deadlines can leave little time to understand code, and that overreliance may create problems.
Practical Benefits and Uses of AI
One commenter framed AI use as part of the practical need to deliver products under workplace pressures.
Key facts
- Experience
- The engineer said they had three years of experience.
- AI use
- They said they used AI for tasks ranging from bug fixes to larger coding work.
- Workplace concern
- They said questions from senior colleagues about their code made them anxious because they struggled to explain it.
- Self-described effects
- They reported reduced focus and confidence, and difficulty reading documentation or tracing code manually.
- Suggested approach
- One commenter said they use AI to code and write tests but verify the code before submitting it.
- Reported risk
- A commenter described bugs and regressions at four companies where, they said, staff did not understand the code they were writing.
Quotes
The software engineer
A software engineer with three years of experience who described relying on AI.
“I feel completely useless, fraudulent, and terrifyingly replaceable. If you took away AI tomorrow, I feel like I'd code at the level of a first-semester freshman. Am I cooked? Has anyone else dug themselves into this hole and actually managed to climb back out?”
NDTV
“Started happening to me also. I scaled back the AI usage. I let it code and write tests, but I verify the code before pushing and creating a PR. Dev testing is always manual with debugging, so I know what the code does.”
NDTV
Sources
Related news
Sridhar Vembu Warns Engineers Against Letting AI Replace Understanding
Sridhar Vembu Urges AI Use Without Losing Technical Understanding
AI Coding Boom Exposes Production Risks and Technical Debt
MIT Study Links ChatGPT Essay Use to Cognitive Debt
Bengaluru Techie Says AI Made 15 Years Experience Feel Obsolete
Why All-Day AI Use Can Cause Mental Exhaustion
Swami Sarvapriyananda Says Anthropic Invited Him for Claude Ethics Discussions
Anthropic’s AI Consciousness Debate Reaches Religious Leaders and Critics
Anthropic Considers November IPO Amid Valuation and Loss Concerns
Anthropic Consults Religious Scholars About AI Morality and Consciousness
