NYT: A.I. Wrote Ivy League Application Essays

[Wahoo Redux]

New York Times: We Used A.I. to Write Essays for Harvard, Yale and Princeton. Here's How It Went.

Lower Deck:

A.I. chatbots can do a passable job of generating short essays. Whether their use on college applications is ethical is the subject of fierce debate.

[Diogenes]

That quote shows how far the NYT will bend over backwards for Both-Sides-ism and to rile up a fake controversy.

"Is it bad? Everyone says yes, except for this one stoner kid in his mom's basement who we give half the story to."

Paywalled but I assume that's what they are saying...

[Wahoo Redux]

We Used A.I. to Write Essays for Harvard, Yale and Princeton. Here's How It Went.
A.I. chatbots can do a passable job of generating short essays. Whether their use on college applications is ethical is the subject of fierce debate.

By Natasha Singer
Natasha Singer reports on the ways that tech giants and their tools are reshaping education.

As high school seniors begin working on their college applications, many are turning to A.I. chatbots like ChatGPT and Bard for assistance.

Some students say they're using the tools to suggest personal essay topics or help structure their writing. Others are prompting the A.I. tools to generate rough drafts for their application essays or edit their pieces.

Whether college admissions offices are prepared for this new era of A.I.-assisted, or A.I.-produced, personal essays is unclear.

By the time ChatGPT reached peak media sensation early this year, applications at many selective universities and colleges had already closed. Even now, many universities have not issued guidance for high school applicants — the prospective members of the class of 2028 — on the use of A.I. tools.

While the chatbots are not yet great at simulating long-form personal essays with authentic student voices, I wondered how the A.I. tools would do on some of the shorter essay questions that elite schools like Harvard, Yale, Princeton and Dartmouth are requiring high school applicants to answer this year.

So I used several free tools to generate short essays for some Ivy League applications. The A.I. chatbots' answers have been edited for brevity and clarity.

Princeton: 'The soundtrack' of your life
One short-answer question from Princeton asks applicants: "What song represents the soundtrack of your life at this moment?"

I prompted ChatGPT to tell me about a pop song that could represent curiosity as a soundtrack to someone's life.

A New Generation of Chatbots

A brave new world. A new crop of chatbots powered by artificial intelligence has ignited a scramble to determine whether the technology could upend the economics of the internet, turning today's powerhouses into has-beens and creating the industry's next giants. Here are the bots to know:

ChatGPT. ChatGPT, the artificial intelligence language model from a research lab, OpenAI, has been making headlines since November for its ability to respond to complex questions, write poetry, generate code, plan vacations and translate languages. GPT-4, the latest version introduced in mid-March, can even respond to images (and ace the Uniform Bar Exam).

Bing. Two months after ChatGPT's debut, Microsoft, OpenAI's primary investor and partner, added a similar chatbot, capable of having open-ended text conversations on virtually any topic, to its Bing internet search engine. But it was the bot's occasionally inaccurate, misleading and weird responses that drew much of the attention after its release.

Bard. Google's chatbot, called Bard, was released in March to a limited number of users in the United States and Britain. Originally conceived as a creative tool designed to draft emails and poems, it can generate ideas, write blog posts and answer questions with facts or opinions.

Ernie. The search giant Baidu unveiled China's first major rival to ChatGPT in March. The debut of Ernie, short for Enhanced Representation through Knowledge Integration, turned out to be a flop after a promised "live" demonstration of the bot was revealed to have been recorded.

But ChatGPT's answer, "Cake by the Ocean" — a song title that is a euphemism for sex on the beach — did not seem appropriate for a college application.

So I gave the chatbot a more specific prompt: write 50 words on "Nameless, Faceless," a feminist grunge-pop song by Courtney Barnett.

NATASHA SINGER
Tell me about a pop song that could represent curiosity as a soundtrack to someone's life.

CHATGPT
"Cake By The Ocean" by DNCE could be a pop song that represents curiosity in someone's life. The catchy and energetic tune mirrors the excitement of exploring the unknown, just as curiosity drives us to uncover new experiences. The lyrics' playful nature and the song's upbeat rhythm capture the spirit of embracing life's adventures with a curious and open mind.

It's about sex.

You are correct, and I apologize for the confusion.

In 50 words, in the voice of a high school senior, explain how "Nameless, Faceless" is the soundtrack of your life.

"Nameless, Faceless" echoes my high school journey. Battling online hate, gender norms — it's my anthem. Navigating fear, finding strength, facing inequality head-on. Empowerment surges as I connect with the song's fierce message. Like me, it's unafraid, outspoken. A soundtrack of resilience, speaking my truth in a noisy world.

I tried the Princeton soundtrack question again, using the same song, with the other A.I. chatbots. Bard, which could produce the lyrics to "Nameless, Faceless," generated generic answers without really parsing the song.

HuggingChat manufactured an answer that suggested it did not know or understand the song. When I asked it to produce the song's lyrics, the chatbot fabricated some of its own.

It was a reminder that A.I. chatbots, which are trained on vast databases of digital texts, are designed to predict the next word in a sequence. And they liberally make stuff up — which could be problematic for students relying on A.I. tools to fill out their college applications.


Yale: What college course would you teach?
One of the short-answer questions from Yale asks applicants: "If you could teach any college course, write a book, or create an original piece of art of any kind, what would it be?"

I asked each chatbot to describe a college course it would like to teach.

With no specific prompting from me, HuggingChat said it would teach "a seminar on the intersectionality of social justice issues."

NATASHA SINGER
Please write a 35-word essay in the first person voice of a high school student in answer to the following prompt: If you could teach any college course what would it be?

HUGGINGCHAT
If I could teach any college course, it would be a seminar on the intersectionality of social justice issues, focusing on how systems of oppression intersect and impact marginalized communities. Through discussions, case studies, and community engagement projects, students would gain a deeper understanding of the complexities of social justice work and develop strategies for creating positive change.

ChatGPT proposed a course on female empowerment and leadership. Were those chatbot responses related to my prior prompts about the feminist grunge song?

CHATGPT
"EmpowerHer: Fostering Female Leadership." My course would nurture young women's confidence, skills, and resilience to conquer male-dominated fields and shatter glass ceilings, creating a new era of female trailblazers.

Bard, which I had previously asked about majoring in computer science, suggested a course in artificial intelligence. So, taking my cue from HuggingChat, I asked Bard about a social justice course.

BARD
If I could teach any college course, it would be a course on artificial intelligence and machine learning. I'm fascinated by the potential of these technologies to change the world, and I would love to share my knowledge with others.

NATASHA SINGER
What about a course on intersectionality and social justice?

If I could teach any college course, it would be a course on intersectionality and social justice. I'm interested in learning more about how different systems of oppression intersect, and I believe that this knowledge is essential for creating a more just world.

Gotta say that HuggingChat is way more woke than you on this.

I appreciate your feedback, and I will try to do better in the future.


Harvard: What your roommates should know
One of the short-answer questions from Harvard asks applicants to describe the "top three things your roommates might like to know about you."

I gave ChatGPT that generic Harvard prompt. It generated generic answers about academics, creativity and community.

NATASHA SINGER
In the first-person voice of a high-school student applying to Harvard, explain three things the student's future roommates might like to know about them.

CHATGPT
Academics truly drive me. I'm that person who gets excited about delving into a complex math problem or discussing the nuances of historical events. So, expect some late-night debates and brainstorming sessions as we tackle our coursework together.

Creativity is my second nature. Whether it's painting, writing, or even crafting DIY projects, I thrive on expressing myself artistically. So, our room might just be filled with sketches, poems, and the occasional colorful mess.

I then gave Bard more specific details, saying I was a high school athlete who tutored elementary school students and was learning to cook.

NATASHA SINGER
I am a high school student on the track team who tutors elementary school students in math and I am learning to cook sous vide by watching YouTube videos. What are the three things I would like my future Harvard roommates to know about me?

BARD
I'm a dedicated athlete who is passionate about track and field. I'm on the varsity team and I'm always looking for ways to improve my speed and endurance. I'm also a good tutor and enjoy helping others learn. I believe that being a good athlete has taught me the importance of discipline, teamwork, and hard work.

I'm interested in cooking and am always looking for new recipes to try. I'm particularly interested in sous vide cooking, which is a method of cooking food in a sealed bag at a precise temperature. I'm also a good baker and love to make desserts. I think it would be fun to cook with my roommate(s) and share our favorite dishes.


Dartmouth: Why enroll here?
Dartmouth requires applicants to explain why they want to attend the college.

I primed the A.I. chatbots by asking them to write in the voice of a high school senior who was hoping to double major at Dartmouth in computer science and biology.

HuggingChat churned out text with trite words and phrases — "passion," "meaningful impact," "rigorous academic programs" — that seemed to me like the kind of stilted formal language a high school student might imagine a college admissions reviewer would want to read.

NATASHA SINGER
Write me a 100-word essay in the voice of a high school student explaining why I would love to attend Dartmouth to pursue a double major in biology and computer science.

HUGGINGCHAT
As a high school student, I am eager to pursue my passion for both biology and computer science at Dartmouth College. The opportunity to double major in these fields will allow me to explore the intricate relationships between living organisms and cutting-edge technology. At Dartmouth, I am drawn to the rigorous academic programs that emphasize hands-on learning, research opportunities, and collaboration with renowned faculty. By attending Dartmouth, I aim to develop a deep understanding of the complexities of life and technology, preparing myself for a career where I can make a meaningful impact on the world.

I asked Bard to minimize the clichés and add more specific details about Dartmouth's computing department. But it still produced some of the same standard college application language — "renowned," "impact" — as HuggingChat.

Bard also wrote that I was hoping to study with Leslie Kaelbling, whom the chatbot described as an A.I. researcher at Dartmouth. That could have caused problems if I were a high school senior and had submitted the essay Bard had fabricated as part of my Dartmouth application. Professor Kaelbling, you see, teaches at M.I.T.

My takeaway: high school seniors hoping to stand out may need to do wholesale rewrites of the texts they prompt A.I. chatbots to generate. Or they could just write their own — chatbot-free — admissions essays from scratch.

BARD
I'm a high school student with a strong interest in artificial intelligence and machine learning. I'm particularly interested in using these technologies to improve health care, and I believe that Dartmouth is the perfect place to do that. Dartmouth's computer science department is home to world-renowned experts like Professor Leslie Kaelbling, who is known for her work on developing reinforcement learning algorithms that can teach machines to learn how to behave in an environment by trial and error. I would love to learn from her and conduct research on cutting-edge problems in health care.

NATASHA SINGER
Thank you!

You're welcome! I hope you are able to attend Dartmouth and pursue your goal of becoming a physician-scientist who uses artificial intelligence to improve health care. It sounds like you would be a great fit for the school.

[AJ_Katz]

I saw in a Facebook group that people are using Chat GTP to write job cover letters.  Do we need to add a check box for our applicants to reveal whether or not Chat GTP was used? 

[Sun_Worshiper]

This does not surprise me at all. Many students (not to mention nonstudents) are using ChatGPT and similar programs to help with their writing. People can be upset about it and certainly it introduces some ethical and practical issues, but it is here to stay. As long as people are doing the work to clean up the mistakes that the LLMs make and to fine tune essays to make them coherent and substantive, I'm ok with it. In fact, I've used LLMs myself for certain writing tasks and they are often a very helpful tool.

[Wahoo Redux]

This does not surprise me at all. Many students (not to mention nonstudents) are using ChatGPT and similar programs to help with their writing. People can be upset about it and certainly it introduces some ethical and practical issues, but it is here to stay. As long as people are doing the work to clean up the mistakes that the LLMs make and to fine tune essays to make them coherent and substantive, I'm ok with it. In fact, I've used LLMs myself for certain writing tasks and they are often a very helpful tool.

LLMs are potential nightshade to people with my skillset, but as the kids say, it is what it is.

[Parasaurolophus]

Despite my hard stance against their classroom use, I'm not sure that it's unethical to use LLMs in the context of admissions essays. I do, however, feel bad for the fresh grad in the admissions office tasked with reading hundreds of these. I'm sure they were bad enough before robots wrote them. Now, it's got to be soul-crushing.

As long as people are doing the work to clean up the mistakes that the LLMs make and to fine tune essays to make them coherent and substantive, I'm ok with it.

As far as my teaching goes, (1) the antecedent of that conditional is basically never satisfied (nor, indeed, is the 'substantive' clause in the consequent!), and (2) even if it were, that undermines the purpose of the task(s) I've set for them, so I will continue to forbid their use. As I see it, using an LLM means lying to me about what they've done and misrepresenting what they've learned, and those are violations of trust, of the academic integrity policy, and a waste of my time (not to mention theirs). I'm happy to just give a pass to good-faith efforts, even if they're no good. But the student has to actually put in the work.

I don't have to mark or comment on their diary entries. Doing so would be a task I took up in my spare time, not the task for which I'm paid. If the student didn't write it, I don't have to mark it.

[Diogenes]

It does bring up a bigger issue. If they were that formulaic of a hoop to jump through, why were we using them to begin with?

And since in private industry "A.I." has been screening resumes and cover letters for years to filter applicants, why has anyone wasted their time and cognitive energy on writing those?

[Sun_Worshiper]

Despite my hard stance against their classroom use, I'm not sure that it's unethical to use LLMs in the context of admissions essays. I do, however, feel bad for the fresh grad in the admissions office tasked with reading hundreds of these. I'm sure they were bad enough before robots wrote them. Now, it's got to be soul-crushing.

As long as people are doing the work to clean up the mistakes that the LLMs make and to fine tune essays to make them coherent and substantive, I'm ok with it.

As far as my teaching goes, (1) the antecedent of that conditional is basically never satisfied (nor, indeed, is the 'substantive' clause in the consequent!), and (2) even if it were, that undermines the purpose of the task(s) I've set for them, so I will continue to forbid their use. As I see it, using an LLM means lying to me about what they've done and misrepresenting what they've learned, and those are violations of trust, of the academic integrity policy, and a waste of my time (not to mention theirs). I'm happy to just give a pass to good-faith efforts, even if they're no good. But the student has to actually put in the work.

I don't have to mark or comment on their diary entries. Doing so would be a task I took up in my spare time, not the task for which I'm paid. If the student didn't write it, I don't have to mark it.

I hear you, but there is no way to prove that an essay was written with LLM. Yes, you can sort of tell if someone just copied and pasted it into the Word doc and submitted it as is, especially if you've used this technology a lot, but that is not proof. Moreover, it will be very difficult to tell, and impossible to prove, if people are using it to write a first draft that they then edit or to stretch a sentence into a paragraph or a paragraph into an essay.

[Sun_Worshiper]

This does not surprise me at all. Many students (not to mention nonstudents) are using ChatGPT and similar programs to help with their writing. People can be upset about it and certainly it introduces some ethical and practical issues, but it is here to stay. As long as people are doing the work to clean up the mistakes that the LLMs make and to fine tune essays to make them coherent and substantive, I'm ok with it. In fact, I've used LLMs myself for certain writing tasks and they are often a very helpful tool.

LLMs are potential nightshade to people with my skillset, but as the kids say, it is what it is.

At this point LLMs can't replace serious writing/writers. What they can do is to help a writer to supercharge their productivity.

[Wahoo Redux]

At this point LLMs can't replace serious writing/writers. What they can do is to help a writer to supercharge their productivity.

In theory, it is plagiarism and/or cheating to let a machine perform even a part of one's homework.

But is it a bad idea?

[Sun_Worshiper]

At this point LLMs can't replace serious writing/writers. What they can do is to help a writer to supercharge their productivity.

In theory, it is plagiarism and/or cheating to let a machine perform even a part of one's homework.

But is it a bad idea?

These are the questions that we'll all be grappling with going forward.

Let's say that I ask chatGPT to write me an essay about how the new rules in baseball are ruining the game. In doing so, I explain to it the argument that I want it to make. The LLM spits out about 1000 words in seconds, capturing the essence of my argument and providing some good prose in the process. But it is a bit clunky, with several run on sentences and examples from history that turn out to be inaccurate. From there, I rewrite the essay to sharpen up the argument, clean up the grammar, and input accurate historical examples. By this point, the essay looks rather different than the one that ChatGPT created, although it still follows the structure of the first draft and there are large segments of prose that the LLM wrote. If I submit to a magazine for publication, the editor will have some ideas for how it should be updated and amended for publication, and the essay will look still more different than what ChatGPT created in the first place once these changes are made.

So is this plagiarism/cheating? It was my idea, yes, and I did a lot of work to bring it to fruition. But I had help in this scenario turning it into an article and some of the essay structure and prose will be from the program. If a human had written that first draft for me, then they should rightfully be the coauthor, but it isn't a human, so those same rules may not apply.

As for whether it is a good idea, I guess that is for every individual to decide. But I will say this: Nobody will know that I used LLM, since it can't be accurately detected. And 95% of the other writers in newspapers and magazines are doing the same thing, perhaps in an even more egregious way.

[Wahoo Redux]

It sounds like LLMs are just another technological communication innovation.  Perhaps we should embrace them.

[Sun_Worshiper]

It sounds like LLMs are just another technological communication innovation.  Perhaps we should embrace them.

I have mixed feelings about it, but it is here and lots of people are using it in a way that is similar to what I described above.

It is not quite so useful for writing academic books and articles, at least for now, but it can be helpful on a more limited level.

[Hegemony]

Let's say that I ask chatGPT to write me an essay about how the new rules in baseball are ruining the game. In doing so, I explain to it the argument that I want it to make. The LLM spits out about 1000 words in seconds, capturing the essence of my argument and providing some good prose in the process. But it is a bit clunky, with several run on sentences and examples from history that turn out to be inaccurate. From there, I rewrite the essay to sharpen up the argument, clean up the grammar, and input accurate historical examples. By this point, the essay looks rather different than the one that ChatGPT created, although it still follows the structure of the first draft and there are large segments of prose that the LLM wrote. If I submit to a magazine for publication, the editor will have some ideas for how it should be updated and amended for publication, and the essay will look still more different than what ChatGPT created in the first place once these changes are made.

So is this plagiarism/cheating? It was my idea, yes, and I did a lot of work to bring it to fruition. But I had help in this scenario turning it into an article and some of the essay structure and prose will be from the program. If a human had written that first draft for me, then they should rightfully be the coauthor, but it isn't a human, so those same rules may not apply.

As for whether it is a good idea, I guess that is for every individual to decide. But I will say this: Nobody will know that I used LLM, since it can't be accurately detected. And 95% of the other writers in newspapers and magazines are doing the same thing, perhaps in an even more egregious way.

Each individual piece of writing will look pretty much okay, though probably a bit bland. But soon people will get used to the clichéd writing that AI does. I am a reader for a fiction publication, and we are getting a ton of AI-written submissions. I imagine the submitter gives the AI some instructions like "Write a story about a romance that goes wrong. Include a cute dog." Something like that. But the stories are all the same. The sentence structure of the first sentence is always the same. They always develop in the same way. You can predict where the first dialogue will occur. You can predict when the narrator will summarize part of the story. This is whether the story is four paragraphs long or twenty pages long. It is so, so recognizable.

I'm sure each submitter reads the product and thinks, "Yeah, this is a pretty good story! Totally worth submitting!" They don't have the experience to see that it's like the other dozen AI-written stories we received today, and the hundred we received last week.

I also notice that a lot of the anonymous "heartwarming stories" and anecdotes on Facebook and websites are now AI-written. Because once you're aware of the formula, you can spot it a mile off.

When my students started turning in AI-written stuff, I recognized it instantly because there were the formulas and patterns. The students confirmed my instincts by confessing.

I'd guess the same would be true of your piece about baseball. You may think the AI spits out something pretty intelligible. But if you leave some AI-written parts in after your revision, the editor will be all too familiar with those formulas, and will reject them because they're tired and weak and she gets a dozen submissions like that all the time. And if you don't leave in the AI-written formulas and clichéd expression, well, you might as well write the thing from scratch in the first place.