I Tried Building a Real App with AI. It Took a Year (alexhyett.com)

Article summary

The author attempted to build a habit-tracking app with AI assistance, but it took a year to complete due to various issues with the AI-generated code. The author had to manually refactor and fix the code, and also deal with changes in iOS. The app is now available on the App Store, but the author notes that using AI for development has its limitations. The author reflects on the role of AI in software development and the importance of human skills and judgment.

Main themes

• AI-assisted development
• Software maintenance
• Code quality
• App development
• iOS development
• AI limitations

What commenters say

• Using AI for development can speed up certain tasks, but it is not a replacement for human judgment and skills.
• The quality of AI-generated code is not always reliable and may require significant manual refactoring.
• Some argue that AI-generated code is not a problem as long as it works, while others believe that code quality is still important for maintainability and scalability.
• The use of AI in development may lead to a loss of human skills and knowledge, making it difficult to maintain and update software in the long run.
• There is a need for a balanced approach to AI-assisted development, where AI is used to augment human capabilities rather than replace them.
• The concept of 'lifetime' licenses and subscriptions is becoming increasingly questionable in the context of rapidly evolving software and app stores.
• Some developers are using AI to generate code, but this approach can lead to a lack of planning and a 'wildly swinging' development process.
• The importance of code quality and maintainability is still a topic of debate, with some arguing that it is less important with the use of rigorous testing and AI-generated code.