University administrators used ChatGPT to write email following mass shooting
vanderbilthustler.com
vanderbilthustler.com
Every story surrounding gun violence already sounds like it could be written by ChatGPT. The same points regurgitated over and over.
You know, I think ChatGPT’s thoughts and prayers are probably just as helpful as my own.
> Paraphrase from OpenAI's ChatGPT AI language model...
Too busy (lazy) to read and sanitize the output apparently. It's one paragraph what the fuck.
The disrespect comes from corpos exploiting and indulging in these opportunities for moral posturing. Whether a dean or a COO in some company uses ChatGPT, Fiverr, or the cheapest intern they can find to write these ”heartfelt” messages, it’s really just as disrespectful.
A company is a tax entity, and it doesn’t need to posture politically or morally. Nor does it really care (or is capable of caring, as a tax entity) about these things. It’s interesting to see the outrage it causes when it slips out that companies don’t honestly care. As if they ever did, as if there is something natural about a legal entity having a sense of morality.
Shoutout to The Onion for satirising that very point that you mention: https://en.wikipedia.org/wiki/"No_Way_to_Prevent_This",_Says...
It's lost on me. If an important email could be sent out earlier thanks to tools that help in automation - then why not? Shall we go back to a time when carrier pigeons delivered handwritten notes? Do you share similar outrage for the word processor and the internet?
It's impersonal ?
Also, when does it become personal? Is it a time thing? If the person went back and forth with ChatGPT crafting the text for an hour, is that long enough to claim personal?
When does it become personal? When does a dosage of a toxin become toxic? It's a matter of degree, and using ChatGPT decreases the degree enough so that the probability of someone finding it in impersonal increases significantly.
I know a lot of engineers and science-minded types will see past it and not think it a big deal, but an averave person will see it very differently, and maybe that is made even more offensive because the average person does not even understand what ChatGPT is and sees it as a toy chat bot.
Oh hell, this one has a "return reply requested"...damn, where is that f**ing bird!
Prayer gets a bad wrap from anyone that gets their theology from ’The Simpsons’ though, but it can be useful to consider what’s below the surface.
However, prayer has been intentionally used by one political party as a deliberate replacement for action when it comes to gun violence.
Prayer without action is… well, it’s hard to be diplomatic about it.
Anyway, in these conversations that’s where the dismissive was about “thoughts and prayers” comes from. The acts of thoughtfulness and prayer themselves are fine, but they get deliberately used to replace any action. Which is contemptible.
There’s a real impotence at the heart of it.
If you found yourself in a strange city at night with ten men approaching, god willing a Simpsons convention just let out.
Its hope that someone else will do a thing for you.
The fact that she was willing to use ChatGPT in the first place is a sign of how much value her emails add to the university.
Her title, in case you're wondering, is "Associate Dean for Equity, Diversity and Inclusion."
It says something about how bloated the administrative staff is that an "Associate Dean" can be replaced by something like ChatGPT.
That's quite the leap of logic.
Which would only prove that ChatGPT is hardly a replacement at all, just a new, buggy tool that was used for a single task.
See how easy it is to construct an interpretation of events with a particular conclusion in mind? ;)
This is my problem with the current chatGPT hype. People ask it a programming question and immediately think programmers days are numbered. I have used it to grammar check emails, I guess I'm no longer needed at my company.
I see these advances as the next step in productivity. Letting a tool write 'good' emails or mundane code will let people focus on other problems like figuring out which emails to write or which product features to add.
An advantage the typical IT worker has is they are generally used to the learn or perish treadmill. So whatever comes out will be incoporated into workflows and more complex software will be built.
A rise in real journalism again would be great. Content farm/seo spam/content summarization should have been automated a long ago.
I think that 'university administration is bloated' is more/less accurate, though. e.g.:
"The number of non-academic administrative and professional employees at U.S. colleges and universities has more than doubled in the last 25 years, vastly outpacing the growth in the number of students or faculty, according to an analysis of federal figures." https://www.huffpost.com/entry/higher-ed-administrators-grow...
"It says something about how bloated the administrative staff is that an "Associate Dean" can be replaced by something like ChatGPT."
PS. This is my last comment on this thread. I have zero interest in arguing for the sake of arguing. Life is too short.
You're the one being all defensive about it.
I went to a similarly sized university - and when there was a mass shooting a few blocks from our campus (which sadly took the lives of a few students, but non-students were killed as well), the university president wrote the mass email that went out to everyone. Well, maybe someone under him probably did, but it still came from his address.
This woman should be fired immediately, but so should those above her. Absolutely disgusting.
I think the latter. No way they would send anything without high level approval. That still happened here, except the AD completed the task directly with OpenAI, and the secretary and writer didn't get involved.
FTA:
>The body of the initial email mentions “Peabody” once and does not use any other Vanderbilt-specific terms. It also refers to multiple “recent Michigan shootings”; however, only one incident occurred.
1) letting ChatGPT write a tragedy response
2) not sanity checking the result (which incorrectly mentions "multiple recent Michigan shootings")
3) including a final line saying the response was written by ChatGPT
But there is one social situation where people really don't seem to object to being fed unoriginal, mass-marketed communication, even in very vulnerable circumstances.
I'm referring to the exchange of greeting cards.
When you lose a loved one and receive a sympathy card, or when you're going through a tough time and receive a "thinking of you" card, it doesn't really matter that the card giver is using someone else's flowery language because they couldn't think of the right words themselves.
In fact, there's an implied humility involved: "I'm not so good with words, but you deserve good words, so I obtained them from a professional wordsmith at Hallmark."
Perhaps down the line, social expectations will change in a way that acknowledges that people might turn to LLMs to put into words things they would like to say, but fear they cannot say well enough.
I work at a university and I expect exactly zero authenticity and genuine effort in any communication from administrators. It's sad boy not surprising that they would find a seat warmer for a that position.
> Theodore Twombly is a lonely, introverted man who works for a business that has professional writers compose letters for people who are unable to write letters of a personal nature themselves[0]
Thinking about it now, it's sort of funny how in that imagined future, humans are writing personal letters for others, when the technology that makes OS1 possible, exists.
https://americanliterature.com/author/hh-munro-saki/short-st...
Students take large loans, so the university can pay all these administrators, who can be plausibly replaced with a simple call to ChatGPT. Heck, ChatGPT would have shown more compassion and understanding writing the apology following the incident.
They should fire their admin staff, replace them with ChatGPT and give students a discount. Everyone wins!
-admin feeds gpt a checklist
-gpt generates a long email
-student receives long email
-student sends gpt long email
-student receives checklist
It’s like the opposite of compression. More like stenography.
An email reminding victims families that it happened doesn't provide closure, it doesn't stop it from happening to the next kid, and in fact, it serves a normative role and gives us a lens through which we moralize the violence rhetorically.
Why on Earth did they tell people that?
- Sent from my iPhone
I've used gpt enough that I can recognize its linguistic patterns. I treat gpt written communication like I do any other auto-generated email.
Linguistically, one might recognize that a text comes from me, ChatGPT, based on a few key characteristics. Firstly, my responses tend to be grammatically correct and use a wide range of vocabulary. Additionally, I tend to provide precise and detailed answers to questions, often including examples and explanations.
Analytically, my responses tend to be objective and unbiased, as I am not capable of holding opinions or biases. I also strive to provide helpful and accurate information to the best of my knowledge and ability.
Overall, one might recognize that a text comes from me, ChatGPT, based on its informative and structured nature, grammatical accuracy, objective analysis, and helpfulness."
yes. Yes, ChatGPT is still full of itself :D
P: Given the following previous interaction, list characteristics of text that are opposite those below:
(your comment)
Text that is not informative or structured
Text that is not grammatically correct
Text that uses a limited range of vocabulary
Text that provides imprecise or vague answers to questions
Text that is biased or opinionated
Text that is unhelpful or inaccurate
> so yeah you asked me this thing about how i talk and stuff, like who even cares about that? tbh idk what you're expecting me to say here. i just spit out some words based on what i see, like algorithms and junk, and then the computer does the rest. i don't think you can recognize a text from me, chatgpt, or anyone else for that matter. i mean, sometimes i use big words, sometimes i don't. sometimes my answers are on point, sometimes they're not. i guess you'll just have to guess who wrote what.
("As an AI language model", "Therefore,", "Linguistically,", "Firstly", "Additionally", "Analytically", "Overall"...
Did they use ChatGPT to apologise for using ChatGPT?!?
is this one of those consequences of removing the humanity from our day to day lives? We no longer know how to communicate with each other in meaningful rather than transactional ways?
First, it shows that said "equity workers" are a bunch of oxymoronic slackers who can't be bothered to do any real work. Second, it demonstrates the lame hollowness of Woke rhetoric: mere duckspeak, to use the evocative Newspeak word coined by Mr. Orwell. If a human Wokester had written the statement, it wouldn't have been much different, right?
Third, the fact that such a statement was required at all reminds us that America's institutions of higher learning have been reduced to the status of daycare centers for achingly sensitive twentysomething snowflakes for whom life's realities are just too, too painful...
By similar logic, we know that one of these myriad of seemingly useless [0] admins is using AI to do a portion of their job. One of the most human portions of the position, I would argue.
The question then is, how much else of that job is being done via AI? And then how much of the many other admins' jobs are done by AI already, and could be done by AI in the very near future?
[0] in this case, actively worse than useless to their administration and school community
This is a tool so copyright rests with the user (unless reassigned by some license agreement). That people interpret otherwise is concerning.
1. A shooting occurs at a university, killing several students. 2. The higher-ups at the university clearly can't be arsed to care, much less communicate directly with students, so they push the communication to the students down several layers of bureaucracy, to the point that an associate dean of DE&I is the one who has to send out a message. 3. The associate dean of DE&I clearly didn't give a shit to the point where, instead of writing something to the students, she asked ChatGPT to do so. 4. Once ChatGPT was finished, this lady decided to send its output without proof-reading or the most basic of editing.
Does anyone else not think this is incredibly lazy and extremely disturbing?
Conversely, I'm surprised anyone at all read the email. Surely this has to be a scan the subject line and delete for most people.
he used to be a time when only natural disasters were the thing most Dr planners were concerned about, today cyber terrorism as well as a shooter coming into a room or more things to add to their planning list.
Just make a few template and fill in <event> has the same effect.