Big Telecom Tried AI Coding Tools — Here's What Developers Really Thought
Bosses expected a productivity boom. Developers got a helpful but unreliable assistant.
At a large Swedish telecommunications company, researchers studied a coordinated roll-out of AI-enabled software development tools (AI-devtools), conducting semi-structured interviews with 12 software professionals across three sites. Their findings on use cases, productivity, frustrations, and tool limitations corroborated prior survey work. Beyond that, the interviews surfaced a management-developer expectation gap that maps onto the TAM2 constructs of subjective norm and voluntariness, and showed that participants weigh perceived risk heavily — a factor TAM2 and similar acceptance models do not represent. The paper's conclusion, according to the article: AI-devtools emerge as helpful but fallible assistants whose value is shaped by organizational expectations, system scale, and developers' skills.
- Researchers interviewed 12 developers at a large Swedish telecom company, not test subjects in a lab — real people using AI tools daily at work.
- AI helped most with small, routine coding tasks but struggled on large, complex systems, where mistakes still needed human catching.
- Managers and developers wanted different things from the same rollout, and worry about risk — rarely measured in tech adoption studies — weighed heavily on the developers.
Why It Matters
If your workplace is rolling out AI, expect uneven gains — and bosses who expect a magic jump.