vault backup: 2026-01-05 21:12:17

This commit is contained in:
2026-01-05 21:12:17 +03:30
parent 5c6c672132
commit 6bed3ecad1
@@ -0,0 +1,55 @@
## Use of Generative AI — Guidelines and Requirements
Generative AI tools (such as ChatGPT, GitHub Copilot, Gemini, etc.) are widely available and may be used **responsibly** in this course as **learning assistants**. The goal of assignments and projects is not only to produce a working program, but to demonstrate your **understanding of programming concepts** and your ability to explain and modify your own code.
For this reason, the use of generative AI is **allowed but regulated**.
---
### 1. Acceptable Uses of Generative AI
You may use generative AI in ways that **support learning**, including:
- Explaining C++ concepts (e.g. loops, arrays, pointers, structs)
- Helping you understand compiler or runtime errors
- Suggesting ideas or approaches in **plain language**
- Assisting with debugging **after you have attempted the problem yourself**
- Improving code readability, comments, or naming
---
### 2 Unacceptable Uses of Generative AI
You may **not** use generative AI to:
- Generate the full project or major parts of it
- Design the overall architecture of the program
- Write core logic that you do not fully understand
- Copy and submit AI-generated code without modification
- Replace your own reasoning, problem-solving, or debugging effort
Submitting code that you cannot explain or modify will be treated as **plagiarism**, regardless of whether AI was disclosed.
---
### 3. Best Practices for Using AI Effectively
If you choose to use AI, the following practices are strongly recommended:
- **Try first**: Attempt the problem yourself before asking AI
- **Ask focused questions**:
Instead of “write the code”, ask “why does this loop not work?”
- **Work in small pieces**:
Use AI for specific concepts, not entire features
- **Rewrite in your own way**:
Never submit code exactly as produced by AI
- **Verify and test**:
Always compile, run, and test AI-assisted code
Using AI as a **tutor**, not as a **code generator**, leads to better learning and better results.
---
### 4. Evaluation and Accountability
During grading or oral evaluation, you may be asked to:
- Explain parts of your code
- Justify design decisions
- Make small modifications to your program
Your grade is based on **understanding and ownership**, not on how quickly the program was completed.