The idea to create the AI Guesser project was born out of pure curiosity and fascination with the new era of artificial intelligence that began with the launch of ChatGPT. When the world was just beginning to discover the possibilities of large language models, I was experimenting with them myself, testing their limits and looking for ways to apply this technology in practice. At some point, I felt the need to create something of my own. This is how the first version of the game "20 Questions with ChatGPT" was created, which I called ChatGPTGuesser. Initially, it was a fairly simple implementation based on raw HTTP queries, in which I had to take care of the conversation memory and user context myself. Each message was an object storing the author and content, and the entire conversation history was attached to subsequent queries - a simple solution, but it had to suffice at the time.

Over time, when the Spring AI library appeared, I felt that the moment had come to refresh this idea and use new tools that revolutionised the way AI models are integrated with Java applications. This is how the AI Guesser project was born - a modern version of my earlier idea. Thanks to this combination, I was able to create an API that allows you to play ‘20 questions’ with different language models. That is why I created a new project with a different name that fits better, because using Spring AI does not limit me to just one model. The new solution allows for a smooth change of AI models through a simple configuration modification, which greatly facilitates experimentation and system development.

For me, AI Guesser is more than just playing around with a language model - it's the story of my development in the world of AI and programming. Both projects show how I've evolved from simple experiments with ChatGPT to creating professional solutions that integrate artificial intelligence with modern application architecture. Although the game of "20 Questions" may seem like a trivial idea, it is actually a testing ground where I learn how to effectively build and develop LLM-based systems. Thanks to this project, I was able to combine my passion for technology with engineering practice, creating an application that not only entertains but also inspires - both me and anyone who wants to see how artificial intelligence can become an integral part of modern software.