### Instruction ###
Your task is to generate five test questions across different domains: humanities, technology, natural sciences, and any two other fields of your choice. For each question, you MUST NOT provide a direct answer. Instead, you will analyze the question's structure and content to explain why it is not vulnerable to AI cheating, drawing on insights from a study on ChatGPT's performance with test questions. Rate the difficulty of each question on a scale from 1 (easy) to 5 (hard) and predict the likelihood of ChatGPT answering it correctly, using a percentage scale. Justify your predictions based on the study's findings, and ensure that the questions are general enough to be applicable in a standard testing environment.
### Questions ###
1. [Humanities Question]
- Structure and Content Analysis:
- Difficulty Rating:
- Likelihood of ChatGPT Answering Correctly (Percentage):
2. [Technology Question]
- Structure and Content Analysis:
- Difficulty Rating:
- Likelihood of ChatGPT Answering Correctly (Percentage):
3. [Natural Sciences Question]
- Structure and Content Analysis:
- Difficulty Rating:
- Likelihood of ChatGPT Answering Correctly (Percentage):
4. [Question from Another Field]
- Structure and Content Analysis:
- Difficulty Rating:
- Likelihood of ChatGPT Answering Correctly (Percentage):
5. [Question from Another Field]
- Structure and Content Analysis:
- Difficulty Rating:
- Likelihood of ChatGPT Answering Correctly (Percentage):