Update syllabus with edit to genAI section

jzarnett · Dec 4, 2023 · d3a472c · d3a472c
1 parent ed25101
commit d3a472c
Show file tree

Hide file tree

Showing 2 changed files with 1 addition and 1 deletion.
diff --git a/syllabus/ece459-syllabus.pdf b/syllabus/ece459-syllabus.pdf
diff --git a/syllabus/ece459-syllabus.tex b/syllabus/ece459-syllabus.tex
@@ -135,7 +135,7 @@ \section*{Evaluation}
 You'll have at least two weeks to do each assignment. Trying to place the assignment due dates in the term is hard. We try to avoid FYDP deadlines, not place them on top of midterms, keep them vaguely in sync with the course material covered so far, and finally we're not allowed to make the due date for them during reading week or after the end of term. Oh, and because there are so many students from so many programs, it is effectively impossible to pick a date that works for everyone. Apologies in advance. That's why there are grace days (see below).
 
 \paragraph{Generative AI.} 
-The teaching team is keenly aware of the existence of generative AI (e.g., Github Copilot, ChatGPT) and how they can be used to write and test code. This is not a bad thing; IDEs, autocomplete, libraries, and tools like linters and even Valgrind have existed for a long time and can all help you to write your code in an efficient way. We know that banning them does not work (how could we check that, even?) and in reality (co-op or full time employment), you'll not only use these things, but will probably be less productive than your colleagues if you do not. So -- we trust you to use these things responsibly. What does using it responsibly mean?
+The teaching team is keenly aware of the existence of generative AI (e.g., Github Copilot, ChatGPT) and how such tools can be used to write and test code. This is not a bad thing; IDEs, autocomplete, compilers, libraries, and tools like linters or Valgrind have existed for a long time and can all help you to write your code in an efficient way by doing things for you that at one time you had to do manually. We know that banning them does not work (how could we check that, even?) and in reality (co-op or full time employment), you'll not only use these things, but will probably be less productive than your colleagues if you do not. So -- we trust you to use these things responsibly. What does using it responsibly mean?
 
 Our asks on this are two simple things. The first is that you should acknowledge if you used code given to you by an AI tool, just as you would if you found something on a website somewhere. The second thing is that you should use the tool(s) to help you, not do it all for you. This means you should understand the output it gives you and can verify that it is correct (remember that this is your professional responsibility under the law). Keep in mind that AI models are subject to ``hallucinations'' (making things up that aren't true or aren't real) and errors (bugs, vulnerabilities).