A recent developer survey indicates a growing reliance on AI coding tools despite declining trust in their accuracy. Developers report that while AI assists with code generation, the resulting 'almost right' code often necessitates increased debugging efforts. This suggests a trade-off between efficiency gains from AI assistance and the potential for added workload due to AI-introduced errors. The survey highlights the need for improved reliability and precision in AI coding tools to maintain developer confidence and maximize their value.