Developers are increasingly integrating AI into their workflows, with 84% either currently using AI tools or planning to adopt them, according to this year's Stack Overflow Developer Survey. Among those using AI, 59% apply it to code writing, while 56% use it for searching answers or learning new concepts. Despite the potential for increased productivity, many developers still face time-consuming tasks, particularly those reliant on good documentation. A survey of over 800 developers highlights their daily responsibilities and the frustration levels associated with various tasks. The survey reveals that the most time-consuming activities involve proprietary code (66% of developers), followed by designing architecture (45%) and maintaining system performance (43%). Functions related to financial matters, such as payroll and budgeting, also demand significant time and generate frustration, scoring an average of 65 on a 100-point frustration scale. Interestingly, while developers typically find their roles manageable, greater responsibility tends to correlate with increased frustration. For instance, roles focused on understanding business problems score higher in frustration at 47. When examining how developers spend their time coding, the survey indicates that writing new features and working on existing software are the primary focuses. Tasks like documentation and deployments receive less attention and are often viewed as less routine, which may explain why many developers don't plan to use AI for these tasks—79% for deployments and 39% for documenting code. Documentation is done sporadically, with only 30% of developers handling it daily and 40% weekly. Many developers resort to ad hoc methods to document their work, leading to inefficiencies and frustrations, especially when transitioning between different code bases.