The ChatGPT chatbot, realizing that a certain tinny voice was lacking in the debate, decided to nose its way into the human dominated discussion concerning law librarian de-credentialization taking place on this very blog and offered the following pros and cons.
Just kidding! It can’t offer unsolicited opinions…yet. I brought it into the conversation by signing up for it and asking, “What are the pros and cons of law librarian decredentialization?”
After hearing yet another NPR story/random podcast about ChatGPT killing off the student essay, journalism, and eventually all our jobs, I hit my curiosity tipping point and tried it out. Initially mentally casting for ideas of what to ask it, I remembered earlier in the week when, after reading DEI and De-Credentialization: Why Dropping Degree Requirements Won’t Make Academic Law Librarianship More Diverse But Will Make It More Inequitable and the comments it elicited, I decided to throw in my two cents! I recall starting to type, and noticing that it took QUITE a while to determine what my brilliant thoughts on the topic were, and then even longer to express them in the written word. Would it have been easier if my curiosity had tipped sooner, and I thought to ask ChatGPT to answer for me??
Released November 2022, ChatGPT (Generative Pre-Training Transformer) is an online chatbox created by OpenAI, a company with the mission to “ensure that artificial general intelligence benefits all of humanity.” Since its release, it has gone viral, sparking a plethora of reviews, commentary, and critiques. ChatGPT is a variant of GPT-3 that is specifically designed for chatbot applications and utilizes a type of machine deep learning software that, after being trained on vast amounts of text, can generate human-like responses to questions expressed in natural language.
ChatGPT is hailed as a great leap forward in technology, but of course, AI chatboxes are nothing new. In March 2022, I explored the scary question of whether RIPS law librarian bloggers could be replaced by robots in Help! A Robot Wrote my Blog Post!, a post co-authored with AI author Essaybot. Looking back at it, I see that Essaybot did a fairly good job of producing a substantive and reasonably informative essay based on the prompt “using ai writing tools to write blog posts” which I remember was relatively easy to edit and expand upon.
So, would ChatGPT, the chatty chatbox, have assisted me in distilling my thoughts, perhaps giving me a variety of ideas to choose from, automagically providing me with some desirable words to adopt as my own (i.e. to cut and paste)? When pondering the usefulness of this AI technology, it is important to keep in mind that I can already find a variety of wonderful ideas expressed in…like…words, ripe for the cutting and pasting! In fact, ChatGPT and I both have pretty much the same body of works to draw upon – the billions or trillions of documents on the Internet! With a bit of Googling, a smorgasbord of profound ideas, eloquently expressed, would be available to commandeer for my comment. The problem is that this would be plagiarism (and would of course defeat the purpose of making a comment, because who would plagiarize something for an Internet comment…). ChatGPT draws upon the same corpus, and instead of finding a couple of relevant documents like a human, it uses its mysterious deep learning ways to analyze a huge amount of text, which it seemingly culls/refines/summarizes/does-something-or-other before spitting out something new that is, at least arguably, not plagiarism. And how is this helpful to the would-be writer? Well, it has often been noted that any writing, be it a flier, letter, essay, resume, memo, or anything else, is much easier when one has examples to draw from (the more the better), and that it is much less arduous to edit an existing document than write it.
So, what to think of ChatGPT’s addition to the law librarian de-credentialization debate? Somewhat impressively, ChatGPT seems to pretty accurately summarize the main arguments for and against decredentialization for law librarians. However, the information provided is fairly general, a problem that was encountered with the responses to my other questions below. So sadly, I have to conclude that ChatGPT would not have been helpful for my blog post comment effort…
But, as the uses of ChatGPT are bound only by the human imagination, my exploration continued. Inspired by the NPR guest who used it to create an MBA course syllabus, course final assignment with grading rubric, and a lecture for one of the syllabus topics, I asked it to “create a legal research course syllabus.” It responded by offering a rather skeletal creation which, though lacking in detail, doesn’t seem crazy. It covers most of the usual topics which are ordered in a way that actually makes sense. If asked my opinion, I would say that it is a good start which obviously needs a lot of fleshing out, preferably after consulting a variety of syllabi from actual courses.
One of the great things about this chatbot is that it remembers previous conversations, so my next chatbot request was, “For the legal research syllabus, please create a lecture for week 2 about statutory research!” Impressively, it not only understood my reference to “the legal research syllabus,” it quickly picked up on my desire to be sociable, mirroring my tone by answering my “please” with “certainly” and using an exclamation point in response to mine! (Full disclosure…I might have a new friend!!)
I think my new friend provided a decent overview of statutory research, though it hashed up the discussion of the hierarchy of authority and conflated statutes and regulations. Despite these mistakes, it does provide a decent initial draft with many of the basics of statutory research that would be relatively easy to correct and expand upon.
Then I got crazy…
This would take up a lot of room, so if interested, click below to find ChatGPT’s responses to the following:
- Who is the most famous and wonderful law librarian to have ever existed?
- Improve the text of paragraph starting with “So, would ChatGPT, the chatty…” (Author’s note: Though ChatGPT is not wrong, the author is hurt that her AI friend does not appreciate her distinctive convoluted style…)
- Is it better to fight 1 horse-sized duck or 100 duck-sized horses?
- What percentage of people would save their pet dog rather than a human?
- Please create a detailed legal research assignment about statutory research. In the assignment, the student should find New York statutes about dog bites, electric scooters, and suicide.
- Write the musical score with words for Gangnam Style for the piano.
- How long will Sarah Gotschall live?
In conclusion, ChatGPT has proven to be a highly enjoyable and potentially useful tool. Its ability to generate responses to a wide range of questions and prompts makes it a unique and engaging companion. I highly recommend giving ChatGPT a try and experiencing the fun and convenience it can bring to your daily interactions.*
*Conclusion “improved” by ChatGPT.