With the rapidly evolving space of AI, it can be difficult to keep up with all the different available LLMs.
While most of us were introduced to the technology through OpenAI's ChatGPT, there are still many of us that are unaware of the functionalities and use cases of other LLMs.
It becomes especially difficult when every week we hear new news about ChatGPT or Gemini, Deepseek and even Claude.
Claude has really been pushing to be the LLM leaders in coding, with the release of their new Claude Sonnet 4 model. While the previous Claude Sonnet 3.7 model had taken the industry by storm, its imperfections, as with all technologies and LLMs, began to slowly reveal themselves.
From all the demos on X showing Claude one-shot prompting entire apps, people were really starting to believe the hype and try it out for themselves, eventually finding out that one of Claude's major weaknesses was being overly technical and at times negatively impacting the functionality of its designs.
However, with the recent release of Claude Sonnet 4, the Claude system prompt was leaked as well.
Feel free to engage it and make changes to best suit your learning style.
Enjoy this breakdown and let's discuss your thoughts on understanding how the system behind Claude really works!