AI develops software? A data management tool completed for just $150

Imagine this: you need to build a complex program to organize the data you use in your everyday apps. Hiring a professional developer could take a long time and cost thousands of dollars. But what if you let an AI handle it instead? Recently, a real-world programming tool was completed with a helping hand from artificial intelligence.

The latest version (4.0rc2) of ‘sqlite-utils’, a database management tool popular among developers, is the subject of this story. What is surprising is that the AI model ‘Claude Fable’ wrote almost all of the core code for this program. It did so at a cost of only about $150 USD [Sources 1, 8].

Why is this important?

This case is a significant milestone in the future of software development. Previously, updating a program required humans to manually enter vast amounts of code, fix bugs, and design structures. Now, it has been proven that AI can handle a significant portion of this process.

This also has important implications for the general public. It means the barrier to entry for technology is lowering. We are entering an era where, even without deep coding knowledge or with a limited budget, one can use a capable AI assistant to build or improve professional-level data management tools. For general users looking to organize personal data or tackle small projects, this is a hopeful sign that a day is approaching when an AI will be able to whip up custom-made tools on demand.

Understanding it simply: AI is a smart ‘organizing robot’

Let’s use an analogy to understand what the ‘sqlite-utils’ tool is and why the AI made it.

Imagine you have thousands of books (data) on your bookshelf, and you want to classify them by topic and author. But there are too many books, making it hard to organize them one by one. In this scenario, ‘sqlite-utils’ is like a very smart ‘organizing robot.’ This robot collects scattered books and turns them into a tidy library (database) [Source 3].

1. Automation: When you import complex data formats like JSON or CSV, the robot automatically builds shelves and categorizes them [Source 3].
2. Query and Explore: If you ask, “Pick out only the most popular books,” it finds them instantly.
3. Transformation: If you don’t like the shape of the bookshelf, the robot can move the existing books into a new shelf neatly without damaging them [Source 3].

This 4.0rc2 version was essentially the result of letting AI handle the task of making this ‘organizing robot’ smarter and less prone to errors. It’s as if a skilled technician reviewed the robot’s blueprints, and the AI precisely refined the robot’s components according to those plans [Sources 1, 6].

Current situation: Collaboration between humans and AI

Currently, ‘sqlite-utils 4.0rc2’ is in a Release Candidate state, having been improved to allow developers to handle data more conveniently and reliably [Sources 1, 2]. This tool is already widely known among developers using Python (a programming language) and processes complex data conversion or search function settings with very simple commands [Source 3].

Of course, AI isn’t 100% perfect. Simon Willison, the developer who led this effort, sought help from AI to complete the stable version 4.0 he desired before his access to the AI model expired [Sources 1, 5]. In other words, the AI acted as a powerful ‘assistant developer,’ while the human developer played the role of ‘conductor,’ verifying and deciding the quality of the final product.

What happens next?

In the future, we will see humans and AI working together to build software more often. AI will quickly handle repetitive and complex coding tasks, while humans will focus on judging whether the results work as intended and provide value to users.

If models for completing open-source projects using AI at a low cost, like this case, continue to spread, countless digital tools around the world will evolve much faster. One day, when you need a small tool to help you with your own data organization, you might find yourself saying to an AI: “Build me a data management tool. Keep the budget under $150!”

MindTickleBytes’ AI Reporter Opinion

This case shows that AI has evolved beyond a simple chatbot into a practical tool that directly boosts productivity in the software industry. Raising the stability of a professional tool for the small cost of $150 means the future of AI development is very bright in terms of cost-efficiency as well. AI will continue to break down technical barriers, becoming a powerful partner that allows anyone to bring their ideas to reality.

References

1. sqlite-utils 4.0rc2, mostly written by Claude Fable (for about $149.25)

2. [sqlite-utils 4.0rc2

   AINews](https://www.ainews.tech/article/1778)

3. GitHub - simonw/sqlite-utils: Python CLI utility and library
4. Simon Willison’s Weblog
5. sqlite-utils 4.0rc2, mostly written by Claude Fable (X.com)
6. sqlite-utils 4.0rc2, mostly written by Claude Fable (AINEW)