Artificial Intelligence (A.I. Resources)

Benefits of AI Generated Content

There are many benefits that accompany AI. As this new technology continues to develop, new ways of incorporating AI into everyday life are established and explored.

Benefits

• It can offer solutions to problem-based questions.
• It can sift through massive amounts of information and provide concise summaries. 
• It can analyze text and data, recognizing patterns.
• It can help fill in gaps in knowledge and serve as a jumping off point for new topics.
• It can provide instant feedback 24/7 without the stigma of criticism.
• It can act as a tutor or study partner to test knowledge and instantly get more info when something does not make sense.
• It can organize information, provide outlines, draft emails and messages.
• It can provide language support for global audiences and help overcome language barriers.
• It can simplify complex concepts.
• It can provide a different perspective on topics.
• It can suggest topics to explore AND provide content.
• It can retrieve relevant facts and current information (depending on the platform).
• It can save time by streamlining work. 
• It can help spawn creativity.
• It can admit its mistakes (when challenged) and reject inappropriate requests.

Drawbacks of AI Generated Content

Along with the benefits of AI generated content comes drawbacks. Keep in mind that content generated by AI tools is not considered a primary or original source of information.

Drawbacks

• Rely on AI too much. Not learn or grow
• Difficult to prove that AI was used
• May not recognize what is acceptable or unacceptable without guidance
• A lacks comprehension and context to the information it provides
• Responses are formulaic
• May include incorrect info, biases, and misinterpretations
• It can provide offensive responses unintentionally
• Citations typically contain mistakes or completely made up info
• AI Content may be based on proprietary information and under copyright
• Content behind paywall subscriptions are excluded (JSTOR, Elsevier, SpringerLink)
• None digital experiences are excluded
• Responses may be plagiarized elsewhere. Users may not know info has been plagiarized
• AI can't think critically or independently; it can't compare to the creativity of the human mind
• AI lacks emotional intelligence that humans use to process information
• Generated information may be misused
• Users may not know how to verify information with other sources before including it in their work

Why Students "Cheat"

Cheating has been around forever, and AI is just one more tool that can be exploited. Even with proper guidance and training, students may still choose to "cheat" with AI for various reasons.

Why a student might "cheat" may indicate larger pedagogical issues:

• Different Understanding of Academic Integrity Policies
  Some students may not fully grasp the seriousness of academic integrity violations. However, sometimes course or assignment expectations are not clearly explained. Instructors need to be clear about the use of AI in assignments, tests, and course expectations. Expectations and directions must be clearly stated in syllabi and assignments, and verbal reminders should also be made. 

• Poor Time Management
  High stress levels and academic pressure can lead students to seek shortcuts. Students may cheat when they run out of time to complete assignments or study effectively. Willamette's Academic Support and Success team can help students with time management. Students may not be use to heavier course load or balancing external activities such as collegiate sports with course work. A student's overall semester workload may be too much, but it is possible that there may actually be too much work for a specific course. 

• Course content is not understood
  This is a red flag. Communication between student and instructor broke down somewhere. While students need to be reminded to ask questions when they don't understand, they must also have ample time to consider what they don't know and formulate questions. If one student does not understand something, other students may feel the same way.

• Fear of Failure
  The fear of failing a course or assignment can drive students to cheat in an attempt to secure better grades. Well-intentioned family and friends may place high or unrealistic expectations on a student. They may have been a straight "A" student in high school or feel the need to get high grades for continued education, and feel the additional burden of meeting familial expectations.

• Peer Influence
  The belief that “everyone does it” can normalize cheating, especially when students see their peers engaging in dishonest behavior. Also, what may be acceptable in one course may be prohibited in a different course; a student may think it is okay to use AI the same way as a completely different course.

• Lack of Interest
  Disinterest in a particular subject or assignment may lead students to plagiarize or cheat. Students may need help seeing the value or purpose of work. If the purpose of an assignment is not clear, it may be time to review the assignment and learning outcomes.

• Insecure
  This is a big factor. Students are not confident in their knowledge, skill, or abilities. They recognize deficiencies in their own writing and feel pressure to write at a higher level than they are currently able to do. They may not know the professional terminology because they have not had enough exposure to it yet. Also, students may not feel like they do not belong or fit in, especially when they receive poor grades. It is important for instructors to provide an inclusive learning environment in their classes and for students to feel free to ask questions and learn from their mistakes. For struggling students, it is important to give students the academic support that they need early on. 

Talking Points About Academic Integrity and AI

Professors may have different opinions about what is acceptable use of generative AI tools.

Ultimately, it is up to the instructor to decide if and to what extent AI will be used in a course. Also, students need to check with their instructor prior to using AI tools.

Why is it important to cite sources?

• Citing gives credit to the source creator’s ideas and words and can strengthen your argument or position. 
• Citing the author or source avoids plagiarism. Plagiarism is the act of claiming another person's language, thoughts, ideas, or expressions as though they were one’s own work. What counts is whether you created it or someone/something created it or helped create it for you.

• A citation is like an address for a source; citing tells the reader where you, the writer, found the source.  Reader can visit the original source if needed. 

Considering the points above, researchers should be aware of the following considerations. ChatGPT...

• May generate citations for sources that do not exist.
• May be unable to provide specific sources of information or citations for the information it shares in a response to a query; its responses come from across a vast corpus.  
• The platform is only as good as the information it draws from, and the current ChatGTP 3.5 (free) platform was released in 2021; the free version is already a few years old.  It also may contain contradictory, biased, or false information.

Example: Upon asking ChatGPT to provide an annotated bibliography on the topic of residential care options for adults with autism, it provided a list with four citations and well-worded annotations. Although they all included real journal titles with corresponding volumes and issues, none of the article titles, authors or pagination matched any of the journals they were listed within. These were fabricated citations.

Example: Upon asking ChatGPT to provide an outline for the topic described above, it gave the response below. It is sometimes unable to explain where information came from or refer to specific sources:
"As for my source of information, I am an AI language model, and I generate my responses based on a vast amount of knowledge that I have been trained on."

Further Reading 

"Using AI to Write Scholarly Publications," by Hosseini, Rasmussen, and Resnik. In Accountability in Research. 
This article provides an academic discussion on the issues with using artificial intelligence tools to write scholarly publications.