AI guidelines for the course - Revision history

WikiSysop: /* Enforcement */

2025-01-31T14:21:50Z

Enforcement

← Older revision		Revision as of 16:21, 31 January 2025
Line 26:		Line 26:
	* Exercises made by AI are voided		* Exercises made by AI are voided
	How will exercises be recognized as AI generated?<br>		How will exercises be recognized as AI generated?<br>
	As AI is trained on "the code available on the internet" almost all AI generated code will contain elements not in the curriculum or not yet gone through in the course progression. Any code that contains these elements ~~are~~ therefore voided.<br>		As AI is trained on "the code available on the internet" almost all AI generated code will contain elements not in the curriculum or not yet gone through in the course progression. Any code that contains these elements is therefore voided.<br>
	This may as a side effect hit advanced students who know more python than expected. If you want to use these elements to show your skill, fine. But do ALSO make a solution without the elements. If you can not do that, maybe you need to study the subject some more and think a bit better.		This may as a side effect hit advanced students who know more python than expected. If you want to use these elements to show your skill, fine. But do ALSO make a solution without the elements. If you can not do that, maybe you need to study the subject some more and think a bit better.

WikiSysop at 14:21, 31 January 2025

2025-01-31T14:21:00Z

← Older revision		Revision as of 16:21, 31 January 2025
Line 25:		Line 25:
	* AI is not allowed at exam		* AI is not allowed at exam
	* Exercises made by AI are voided		* Exercises made by AI are voided
	How will exercises be recognized as AI generated? If ~~they~~		How will exercises be recognized as AI generated?<br>
			As AI is trained on "the code available on the internet" almost all AI generated code will contain elements not in the curriculum or not yet gone through in the course progression. Any code that contains these elements are therefore voided.<br>
			This may as a side effect hit advanced students who know more python than expected. If you want to use these elements to show your skill, fine. But do ALSO make a solution without the elements. If you can not do that, maybe you need to study the subject some more and think a bit better.

WikiSysop: /* Enforcement */

2025-01-31T12:36:17Z

Enforcement

← Older revision		Revision as of 14:36, 31 January 2025
Line 25:		Line 25:
	* AI is not allowed at exam		* AI is not allowed at exam
	* Exercises made by AI are voided		* Exercises made by AI are voided
	How will exercises be recognized as AI generated		How will exercises be recognized as AI generated? If they

WikiSysop at 12:27, 31 January 2025

2025-01-31T12:27:57Z

← Older revision		Revision as of 14:27, 31 January 2025
Line 4:		Line 4:
	The purpose of AI in the course is to help you learn - not to solve your problems for you - possibly (likely) in a way you do not understand.<br>		The purpose of AI in the course is to help you learn - not to solve your problems for you - possibly (likely) in a way you do not understand.<br>
	DTU's honour code is to be followed in all DTU activities; courses and exams among them. It has little or no influence if you are employing "good use" of AI, but upon employing "bad use" of AI, you have to cite your source every time.		DTU's honour code is to be followed in all DTU activities; courses and exams among them. It has little or no influence if you are employing "good use" of AI, but upon employing "bad use" of AI, you have to cite your source every time.

			A simple way for yourself to determine if you have used AI wrong is:
			* Can you explain how your solution works?
			* Do you think '''I''' will be happy with the explanation? (a bit subjective)
			* Have functions/methods been used which are not in the curriculum - or later in the curriculum?

	== Good use ==		== Good use ==
	Query the AI for knowledge:		Query the AI for knowledge:
	Asking questions like "how does this function/method work?", "what is the significance of this?", "explain that" "why is this a good/bad idea" are good types of questions. They help you learn.		Asking questions like "how does this function/method work?", "what is the significance of this?", "explain that" "why is this a good/bad idea" are good types of questions. They help you learn.
			* Critical thinking
			* Judgement
			* Analysis

	== Bad use ==		== Bad use ==
	Query the AI for solutions: Asking questions like "how do I do this?", "can you make python code that does this?" simply solves a problem for you - maybe in a way that is not intended in the course progression and you do not really understand, because it is either outside or too early in the curriculum.<br>		Query the AI for solutions: Asking questions like "how do I do this?", "can you make python code that does this?" simply solves a problem for you - maybe in a way that is not intended in the course progression and you do not really understand, because it is either outside or too early in the curriculum.<br>
			* Copy/paste behavior
			* Potential for over-reliance - not developing problem-solving skills

	== Enforcement ==		== Enforcement ==
			* AI is not allowed at exam
			* Exercises made by AI are voided
			How will exercises be recognized as AI generated

WikiSysop at 12:06, 31 January 2025

2025-01-31T12:06:21Z

← Older revision		Revision as of 14:06, 31 January 2025
Line 3:		Line 3:

	The purpose of AI in the course is to help you learn - not to solve your problems for you - possibly (likely) in a way you do not understand.<br>		The purpose of AI in the course is to help you learn - not to solve your problems for you - possibly (likely) in a way you do not understand.<br>
	DTU's honour code is to be followed in all DTU activities; courses and exams among them. It has little or no influence if you are employing "good use" of AI, but employing "bad use" of AI, you have to cite your source every time.		DTU's honour code is to be followed in all DTU activities; courses and exams among them. It has little or no influence if you are employing "good use" of AI, but upon employing "bad use" of AI, you have to cite your source every time.

	== Good use ==		== Good use ==

WikiSysop at 12:05, 31 January 2025

2025-01-31T12:05:06Z

← Older revision		Revision as of 14:05, 31 January 2025
Line 2:		Line 2:
	However, a proper responsible use of AI can to some extent replace a discussion with a fellow human.		However, a proper responsible use of AI can to some extent replace a discussion with a fellow human.

	The purpose of AI in the course is to help you learn - not to solve your problems for you - possibly (likely) in a way you do not understand.		The purpose of AI in the course is to help you learn - not to solve your problems for you - possibly (likely) in a way you do not understand.<br>
			DTU's honour code is to be followed in all DTU activities; courses and exams among them. It has little or no influence if you are employing "good use" of AI, but employing "bad use" of AI, you have to cite your source every time.

	== Good use ==		== Good use ==

WikiSysop: /* Bad use */

2025-01-31T11:09:12Z

Bad use

← Older revision		Revision as of 13:09, 31 January 2025
Line 9:		Line 9:

	== Bad use ==		== Bad use ==
	Query the AI for solutions: Asking questions like "how do I do this?", "can you make python code that does this?" simply solves a problem for you - ~~likely~~ in a way that is not intended in the course progression and you do not really understand, because it is either outside or too early in the curriculum.<br>		Query the AI for solutions: Asking questions like "how do I do this?", "can you make python code that does this?" simply solves a problem for you - maybe in a way that is not intended in the course progression and you do not really understand, because it is either outside or too early in the curriculum.<br>

	== Enforcement ==		== Enforcement ==

WikiSysop at 11:04, 31 January 2025

2025-01-31T11:04:31Z

← Older revision		Revision as of 13:04, 31 January 2025
Line 2:		Line 2:
	However, a proper responsible use of AI can to some extent replace a discussion with a fellow human.		However, a proper responsible use of AI can to some extent replace a discussion with a fellow human.

	~~## Good use ##~~		The purpose of AI in the course is to help you learn - not to solve your problems for you - possibly (likely) in a way you do not understand.

			== Good use ==
			Query the AI for knowledge:
			Asking questions like "how does this function/method work?", "what is the significance of this?", "explain that" "why is this a good/bad idea" are good types of questions. They help you learn.

	## Bad use ##		== Bad use ==
			Query the AI for solutions: Asking questions like "how do I do this?", "can you make python code that does this?" simply solves a problem for you - likely in a way that is not intended in the course progression and you do not really understand, because it is either outside or too early in the curriculum.<br>

			== Enforcement ==
	## Enforcement ##

	~~A single example of above:<br>~~
	~~With the rise of Copilot/ChatGPT answers are easier to get than ever. If you insist on using such tools, be sure you do not use them in a way harmful to your learning.~~
	~~Specifically, ask questions like "how does this work?", "what is the significance of this?" as these questions promote learning.~~
	Do NOT ask questions like "how do I do this?", "can you make python code that does this?", which simply solves a problem for you - likely in a way that is not intended in the course progression and you do not really understand, because it is either outside or too early in the curriculum.

WikiSysop at 10:30, 31 January 2025

2025-01-31T10:30:12Z

← Older revision		Revision as of 12:30, 31 January 2025
Line 1:		Line 1:
	It is strongly preferred that you do '''not''' use AI in the course. The skills you need to acquire are best acquired by your own effort or perhaps in a discussion with a fellow student or TA/teacher.<br>		It is strongly preferred that you do '''not''' use AI in the course. The skills you need to acquire are best acquired by your own effort or perhaps in a discussion with a fellow student or TA/teacher.<br>
	However, a proper responsible use of AI can to some extent replace a discussion with a fellow human.		However, a proper responsible use of AI can to some extent replace a discussion with a fellow human.

			## Good use ##


			## Bad use ##


			## Enforcement ##

	A single example of above:<br>		A single example of above:<br>

WikiSysop at 10:27, 31 January 2025

2025-01-31T10:27:30Z

← Older revision		Revision as of 12:27, 31 January 2025
Line 1:		Line 1:
	It is strongly preferred that you do '''not''' use AI in the course. The skills you need to acquire are best acquired by your own effort or perhaps in a discussion with a fellow student or TA/teacher.		It is strongly preferred that you do '''not''' use AI in the course. The skills you need to acquire are best acquired by your own effort or perhaps in a discussion with a fellow student or TA/teacher.<br>
			However, a proper responsible use of AI can to some extent replace a discussion with a fellow human.

	A single example of above:<br>		A single example of above:<br>
	With the rise of Copilot/ChatGPT answers are easier to get than ever. If you insist on using such tools, be sure you do not use them in a way harmful to your learning.		With the rise of Copilot/ChatGPT answers are easier to get than ever. If you insist on using such tools, be sure you do not use them in a way harmful to your learning.
	Specifically, ask questions like "how does this work?", "what is the significance of this?" as these questions promote learning.		Specifically, ask questions like "how does this work?", "what is the significance of this?" as these questions promote learning.
	Do NOT ask questions like "how do I do this?", "can you make python code that does this?", which simply solves a problem for you - likely in a way that is not intended in the course progression and you do not really understand, because it is either outside or too early in the curriculum.		Do NOT ask questions like "how do I do this?", "can you make python code that does this?", which simply solves a problem for you - likely in a way that is not intended in the course progression and you do not really understand, because it is either outside or too early in the curriculum.