黑料正能量

On this page:

• The evolving situation of 黑料正能量 student use of AI听
• How might we talk with students?听
• Tools and resources听
• Suggestions for syllabus statements听
• Sample syllabus statements听

The evolving situation of 黑料正能量 student use of AI听

It is probable that students in your classes will use generative AI as they produce work for your classes. Our survey of students made it clear how quickly ChatGPT and other tools have become a part of the furniture of many students鈥� minds. They are constantly engaged by languages that normalize the use of generative AI tools in professional and academic contexts.听听

It will not be straightforwardly possible for you simply to 鈥榖an鈥� the use of generative AI tools. Various levels of generative AI are now embedded in relatively neutral tools, including Microsoft 365, Grammarly, Adobe products, and google searching, and in the operating systems and architecture of phones and PCs. Students will not always know that they are using generative AI.听

Tools that promise to identify writing (and other work) produced directly by generative AI tools such as ChatGPT are not reliable. They have been shown to produce false negatives and false positives (); they are particularly unreliable when assessing the writing of those who have learned to write English as adults (). A range of tools exists with the express purpose of making AI-generated writing undetectable ().听

You should not be confident that, as a human, you can tell the difference between human and AI-generated work, unless you know the writing of the individual student well. In an experiment at the University of Reading, AI-produced writing was submitted alongside student-produced work for a psychology exam, and graded blind by examiners who were not aware of the experiment.听 Examiners did not identify the AI-produced work, and AI did better on average in the written examination than human students ().听

Students are concerned about the implications of AI for their learning. They are concerned about other students鈥� cheating; they are worried that reactive decisions will be made, reducing the interest and openness of assignments for their classes, and thus the intrinsic value of their education. They are also looking for guidance from us as they face a world in which the use of generative-AI tools is being normalized.听

How might we talk with students?听

The flood of language about generative AI means that we have to be 鈥� or we have an opportunity to be 鈥� explicit about the value of liberal-arts learning. The use of AI is less likely to pose problems if students understand why and how you are asking them to work, and if they are enthusiastic about that learning. from Stephanie Laggini Fiore at Temple offers some simple, well-structured advice on how to talk to students.听

It is important that you address AI within your approach to learning in general. We teach from very different positions, in very different disciplines. that disciplines, courses, and professors might situate themselves as 鈥榟umanist鈥� (stressing that some uses of AI will impede the development of human capabilities), 鈥榯echnologist鈥� (interested in ways that use of AI can amplify leaning and solve complex problems), or as 鈥榩ost-humanist鈥� (interested in what emerges in the interaction between humanist and technologist approaches).听听听

You might talk with students about AI in the context of more focused discussions of good practice in scholarship in your discipline (attribution of sources, entering scholarly conversation, etc.). That conversation might include, but not be limited to, discussion of academic integrity. 黑料正能量 doesn鈥檛 yet have a general policy on the use of AI, but this note has been added to the Academic Integrity Policy:听听

Note on Artificial Intelligence. Currently, the American University has no general policy on the use of AI (including generative AI tools such as ChatGPT, Sudowrite, Claude, etc) by students in their work. However, many ways of using these tools are clearly in breach of the existing 黑料正能量 Academic integrity pledge, signed by students, which states that 鈥榓ll academic work submitted at 黑料正能量 must be the product of the student鈥檚 own reflection, study and research鈥�, and prohibits 鈥榯he use of any ideas, words, or data 鈥� without properly attributing their source(s)鈥�. Students should refer to the course syllabus for guidelines as to what uses of AI are permitted, and what uses are forbidden, for each particular course.听

You might want to require students to include declarations about their use of AI use as part of work they submit (including oral presentations). This example comes from Betsy Barre at Wake Forest University, and has generously been made available for re-use and modification without acknowledgement:听

听

Artificial Intelligence Disclosure (check boxes)听

I did not use artificial intelligence in creating this paper/presentation听

I did use artificial intelligence in creating this paper, namely ____________ (ChatGPT, Bard, etc.). I used it in the following ways (check which of the following acceptable uses were utilized)听

• Brainstorming help听

• Outlining help听

• Background information听

• Grammar/spelling/punctuation/mechanics help听听

I affirm I did not generate text with artificial intelligence and copy it into my paper.听

听

Adversarial conversations about use of generative AI are unlikely to be successful (it will be difficult to prove that a student who denies unauthorized AI use is lying). We may have to spend more time talking with students about their processes and their understanding than we did in the past, about how they produced their work. We may have to be explicit about how we want students to document their process. Asking probing questions about student work, and expecting students to defend their understanding, may have to be normalized: this is delicate, as students can tend to perceive questions about how they produced their work as accusations of misconduct.听

Tools and resources听

While some faculty feel that a return to hand-written in-class work is the answer to evading AI-produced writing, handwriting is increasingly alien for some students, and imposing it poses accessibility issues. For in-class writing exercises, you may find it useful to have students write using鈥疞ockDown Browser.鈥疶his is a setting in Blackboard that the Academic Technology team can help you to configure. We suggest you practice it with your students on a shorter, low-stakes writing exercise before using it in earnest.You can reach Academic Technology via email at鈥�ultrahelpaup.edu (ultrahelp)ultrahelpaup.edu (@aup.edu)鈥痜or further information and assistance with LockDown Browser.听

Although 鈥楢I Detection鈥� tools are not reliable enough to serve as proof that a student has used generative AI, should you wish to experiment with those webtools, or activate plagiarism checking software鈥�(TurnItIn)鈥痜or your assignments, Academic Technology can assist you with these as well.听

The experts in the writing lab (writinglabaup.edu), the other ARC tutors (arcaup.edu), and the Teaching and Learning Center (tlcaup.edu), can also help you with developing assignments and working with students to encourage good practice.听

Several academic professional bodies (such as the MLA and APA) are beginning to provide guidance on how to cite AI generated materials. See ; or the for examples.听

Suggestions for syllabus statements听

We have gathered some resources on the听 Learning about AI pages. There is a crowdsourced selection of extremely varied syllabus statements collected by ; the offers an interesting model for reflection as syllabi are built; Torrey Trust of Amherst has made available (for a course about education and technology), which lists allowable and non-allowable uses of generative AI, and justifies the permissions and prohibitions. We have found it a clarifying source for reflection, and some of you might want to borrow elements of her statement.听听

If you come up with formulations that you are pleased with, and are willing for other 黑料正能量 faculty to see them, we would be extremely grateful if you are willing to share them , or by emailing us at aiaup.edu. We think that the best statements for 黑料正能量 will emerge within the culture of 黑料正能量.听

However, we recognize that some of you may want a model to work from. Below are three approaches that you are welcome to copy, to cannibalize, to modify, or to use in your own reflective production of statements for your syllabus. We recognize that different disciplines will need to take different approaches 鈥� another good practice, of course, will be to talk to colleagues in your department about their syllabi.听

If you would like to discuss modifications to your syllabus, or ways of talking to students, contact Geoff Gilbert.听

Sample syllabus statements听

Sample 1: When you do not want students to use any generative-AI tools in their courses听

In this course, you are asked to develop your skills in [name those skills 鈥� eg. research, careful analytic and synthetic reading, and written expression]. While use of generative AI may produce artefacts that appear to demonstrate those skills, it will impede your own development as a thinker, researcher, and writer. For this reason, any use of generative AI is forbidden for all exercises in this course. You should expect, as part of the process of reflecting on your learning, that I will regularly ask you to speak or to write about how you built your essays, and you should keep notes and drafts as evidence of that process. For every assignment, you will be asked to sign an 鈥楢rtificial Intelligence Disclosure鈥�. If I believe that you have inadvertently or carelessly included material produced by AI as part of submitted work I will ask you to rewrite and resubmit the assignment. If I believe you have used AI deceitfully, you will fail the assignment or the course, and a report will be filed with the Academic Integrity Office.听

听

Sample 2: When you wish to encourage experiment with specific generative-AI tools听

I expect you to use generative AI [name specific tools] in this class, and to learn how to use AI tools responsibly and ethically. Some assignments will require it, and guidance about the use of tools will be provided in the course materials and in class. However you should be aware of the limitations of ChatGPT and other tools. You will need to develop skills in writing prompts, you will become aware of the biases built into the data with which AI works, and will be skeptical the solutions it provides. You are responsible for errors in the work produced with the aid of the tool, and you are responsible for attributing reliable sources for the information that it provides. Please provide a paragraph at the end of each assignment in which you detail how you used AI (including the prompts you provided, the processes by which you checked the results, and an account of why you believe that AI was an appropriate tool for the exercise). Failure to do so may leave you in breach of Academic Integrity policy.听

听

Sample 3: To acknowledge the new AI environment and help students to negotiate it听

Recent developments in generative artificial intelligence are fascinating, and the proliferation of tools offering to generate texts, code, and images on the basis of large language models is hard to ignore. AI is probably now embedded in the operating system of your phone and computer, and in the basic software that we use for writing, web searches, etc. We will regularly discuss AI in this course 鈥� as an economic, social, and technical phenomenon 鈥� and we will try to become as conscious as we can about the relation between our own thinking and the infrastructures and presuppositions on which AI draws. Each assignment will include detailed instructions as to which uses of AI are allowed, and which are prohibited. As in all courses, you are solely responsible for the contents and form of the work that you produce, and you should expect to be able to document and cite all of the processes involved in the construction of听 work that you submit.听听