I hate ChatGPT

Discussion in 'General Distance Learning Discussions' started by Mac Juli, Jun 9, 2023.

Loading...
  1. Jonathan Whatley

    Jonathan Whatley Well-Known Member

  2. SteveFoerster

    SteveFoerster Resident Gadfly Staff Member

    "Harvard ’21 grad" isn't a sign of humility (stereotypically speaking, of course).
     
    Maniac Craniac likes this.
  3. jonlevy

    jonlevy Active Member

    I'd say at at least 30% or more of my online students (I instruct at 4 universities in two fields) are using ChatGPT for discussion boards and then just inserting random general cites at the bottom. I know this because after 16 years of teaching online discussion board postings are going from barely a paragraph to 3 0r 4 well written paragraphs. The universities are well aware of this issue, the only fix I have seen is supposedly TII can generate a ChatGPT score, but no idea if it is reliable or not. I'd say this is the death of the discussion board as a serious tool for learning and calls into question the weekly written assignments, which have always been copy and paste for the most part anyway.
     
  4. Rich Douglas

    Rich Douglas Well-Known Member

    After spending years writing a doctoral thesis, my examiners spent 1 hour with me. And most of that was regarding recommended additions. In 10 minutes they knew it was my work.

    Perhaps universities will need to re-think how they measure academic accomplishment, ways that are not susceptible to cheating. Having someone actually discuss their work remains tried-and-true.
     
  5. jonlevy

    jonlevy Active Member

    That is a non starter for rubric based online schools with 20 to 35 in a class. The rubric system which ironically seems to please acceditors already assures that even the most feeble answers get at least a C or B. An adjunct earning $2200 a class has no incentive to root out plagiarism besides glancing at a TII score and even then, it has to be fairly flagrant to justify spending an hour or more writing it up.
     
  6. Rich Douglas

    Rich Douglas Well-Known Member

    In the current paradigm, perhaps. But as I implied in my post, that paradigm might need a pretty big shift.
     
    Helpful2013 likes this.
  7. chrisjm18

    chrisjm18 Well-Known Member

    Sounds like SNHU lol
     
    jonlevy likes this.
  8. Pugbelly2

    Pugbelly2 Active Member

    I quit social media about 8 years ago. Today I only have a LinkedIn account (that I don't maintain) because my business marketing folks tell me I have to. I have no FB, Twitter, Snapchat, Instagram, nothing. I felt pretty out of touch at first but it became really liberating. Fight the urge!
     
    chrisjm18 likes this.
  9. Rich Douglas

    Rich Douglas Well-Known Member

    I have an FB only because my wife does. She's barely active on it and I am not at all.

    I'm pretty active on LinkedIn for business purposes. As a consultant, it can be helpful. As a fully-booked consultant, not so much.
     
  10. chrisjm18

    chrisjm18 Well-Known Member

    This summer, I created discussion prompts that I thought would not be AI friendly. However, I am not sure if it will work. Here is a prompt I entered in ChatGPT:

    Discuss a recent case (no older than 2020) where a public official was accused of being corrupt, unethical, or violating their oath of office. An example could be an elected official at any government level or a government agency employee (federal, state, or local), such as judges, police officers, etc. What happened? What damage occurred as a result? What was the outcome (if the case has already been resolved)? Finally, do you think the sanction was appropriate if the case was already resolved? Why or why not? If the case is still in the legal process, what sanction do you think is appropriate? These could include fines, restitution, probation, nontraditional sanctions, and incarceration. Reviewing the sanctions in Chapter 4 (Environmental Crime) may be helpful.

    ChaptGPT

    As of my last knowledge update in September 2021, I do not have access to information on events that occurred after that date. Therefore, I cannot provide information on a case that took place after 2020. However, I can provide an example of a well-known case that occurred before 2020 involving a public official accused of corruption.

    One example that fits the criteria is the case of Rod Blagojevich, the former Governor of Illinois, USA. In December 2008, Blagojevich was arrested on federal corruption charges, including allegations of attempting to "sell" or trade the appointment to the U.S. Senate seat vacated by then-President-elect Barack Obama.

    The damage caused by Blagojevich's alleged actions was significant. Public trust in the state's government was undermined, and it highlighted concerns about political corruption and unethical behavior in the state's political system. The case received widespread media attention and had a negative impact on the reputation of the state of Illinois and its political leadership.

    The legal process spanned several years, and in 2011, Rod Blagojevich was convicted on multiple corruption charges, including wire fraud, attempted extortion, and bribery. He was sentenced to 14 years in federal prison, one of the most severe sentences handed to a former U.S. governor in recent history.
    Regarding the appropriateness of the sanction, opinions may vary. Some may argue that the sentence was too harsh, while others may see it as justified given the severity of the offenses and the need to send a strong message against corruption in public office.

    If a case is still in the legal process, it would be inappropriate for me to speculate on what sanction would be suitable as I lack up-to-date information and the specific details of the case. Determining an appropriate sanction involves considering the severity of the violations, the impact on the public, the extent of harm caused, and the need for deterrence to prevent similar misconduct in the future. The judicial system and relevant legal authorities are responsible for evaluating the evidence and determining the appropriate sanction based on the specific circumstances of each case.
     
    Dustin and Jonathan Whatley like this.
  11. chrisjm18

    chrisjm18 Well-Known Member

    That was obviously an older case that does not meet the requirement. So, I entered the prompt again.

    ChaptGPT

    As an AI language model, I don't have real-time data and my knowledge is up to date only until September 2021. Therefore, I don't have access to recent cases that occurred after that date. To find information on a recent case that fits your criteria, I recommend checking news sources or conducting a search for cases that occurred in 2020 or later involving public officials accused of corruption, unethical behavior, or violating their oath of office. There may be instances involving elected officials at various government levels or government agency employees such as judges, police officers, etc.

    Once you have identified a specific case, I'll be more than happy to discuss the details, including what happened, the damage that occurred, the outcome, and the appropriateness of the sanctions if the case has been resolved. If the case is still in the legal process, I can provide general information about potential sanctions that may be applicable in similar situations, as outlined in Chapter 4 (Environmental Crime) or other relevant sources.
     
  12. Suss

    Suss Active Member

    You can ask your students to post their citations, with links, directly on the discussion board. Then anyone can check them for authenticity, as well as relevance. No more rando citations. That won't fix the problem 100% but it will be a good start.
     
    sanantone likes this.
  13. Johann

    Johann Well-Known Member

    Yes - sounds like a good start. It's like the old "lawyer" joke.
    "What do you call 100 ChatGPT AI's at the bottom of the East River? -- A good start." :)
     
  14. Dustin

    Dustin Well-Known Member

    I'm not sure that would be a good fix. While my PhD course discussion boards require me to cite sources inline in APA format (and I do), an LLM like ChatGPT can also generate text with in-text citations.
     
    chrisjm18 likes this.
  15. chrisjm18

    chrisjm18 Well-Known Member

    Agree. Maybe I'm a bad professor because there's no way I will check the references for each student unless I have my suspicions about a student's work.
     
  16. Johann

    Johann Well-Known Member

    I doubt it. Genuinely bad professors would never entertain any possibility that they're less than stellar. I think you can safely consider - then dismiss that idea.
    Hmm - considerable experience in law enforcement, B.S, M.S. and PhD degrees in Criminal Justice? Your suspicions-and-hunches quotient and your investigative skills are likely very well-developed and accurate by now, I'd say. You're way better- equipped than most to find cheaters... and I'm sure you will. :) They're gonna be sorry!
     
    Last edited: Jul 24, 2023
    SteveFoerster and chrisjm18 like this.
  17. chrisjm18

    chrisjm18 Well-Known Member

    And I do every semester. Last semester I had one of the dumbest cases of academic dishonesty :D

    I usually like to take the pressure off students by not requiring a final exam. Instead, I require a reflection paper (I changed it to a reflection exam for the upcoming semester). The paper is a bit more than a summary of the most impactful aspect(s) of the course. Students are asked to connect with the highlighted content (e.g., how it relates personally - experience, family, future career goals, interests, etc.).

    Anyway, this student decided to submit another student's paper, changing only the name on the title page and the word "attorney" to "lawyer" in the body of the paper. As I was reading, I knew I read this paper before. So, I went back and looked at the ones I had graded. My suspicions were confirmed. I gave both a zero. However, the student who submitted first reached out after submitting the final grades (she went from A to C). She basically told me what happened and sent me screenshots. The other student wanted guidance on how to structure the paper. She sent her paper to the student, who didn't use it as a guide. Instead, she submitted an almost identical paper.

    Having reviewed the screenshots and considered her explanation, I concluded that her intent was not to aid in academic misconduct but she was merely trying to be helpful. I cautioned her against sharing her paper in the future. I suggested she provide general guidance if she wants to help but not send her paper to anyone. So, I submitted a grade change. It gained the necessary approvals, and she earned an "A."

    The other student had an "F." She had the nerve to reach out to me. I dislike academic dishonesty. It's one of the least favorite things I have to handle each semester.
     
    Rachel83az, Dustin and Mac Juli like this.
  18. Dustin

    Dustin Well-Known Member

    Rampant cheating is what led me to end my term at University of the People early. I was shocked that half of the class (or more) were submitting discussion posts they'd found on the internet or completely nonsensical posts since they'd run them through an "article spinner" that replaced a bunch of the words and turned it into a word salad.
     
    Mac Juli and chrisjm18 like this.
  19. SteveFoerster

    SteveFoerster Resident Gadfly Staff Member

    100% agree!

    Agreed!

    [​IMG]
     
    Jonathan Whatley and chrisjm18 like this.
  20. sanantone

    sanantone Well-Known Member

    Can it generate links? I haven't seen that it can generate a DOI or URL to a non-existent source.
     

Share This Page