Podcast: Play in new window | Download | Embed
Subscribe: Apple Podcasts | Google Podcasts | Spotify | Android | Stitcher | TuneIn
Show Summary
In this podcast episode, Linda Abraham discusses the use of ChatGPT in MBA application essays with admissions directors at top MBA programs. The admissions directors generally view ChatGPT as a tool that can be used to enhance the application process, but they emphasize the importance of authenticity and personal reflection in the essays. They caution against relying too heavily on ChatGPT and stress the need for applicants to take ownership of their ideas and experiences. The admissions directors also discuss the potential benefits and limitations of using ChatGPT and suggest ways in which it can be used effectively as a tool. Overall, they encourage applicants to use ChatGPT thoughtfully and responsibly, while still putting in the necessary effort to create thoughtful and authentic applications
Show Notes
ChatGPT roared into our consciousness at the very end of last year, and I had the opportunity to ask several admissions directors what they thought about applicants using it. That’s what we’re going to discuss today.
Welcome to Admissions Straight Talk. [0:46]
Welcome to the 556th episode of Admissions Straight Talk, thanks for tuning in. Before I turn to the subject of today’s show, I have a question for you. Are you ready to apply to your dream MBA programs? Are you competitive at your target schools? Accepted’s MBA admissions quiz can give you a quick reality check. Complete the quiz and you’ll not only get an assessment, but tips on how to improve your qualifications.
Now, if you are a regular listener, you know that during most episodes of Admissions Straight Talk, I interview a guest, frequently an admissions director or dean. I also have many times asked these guests, “What do you think about applicants using ChatGPT or artificial intelligence when writing their application essays?” Today’s episode is a collection of their answers to that question with a little commentary from me, but mostly it’s admissions directors at top MBA programs, sharing what you need to know – the good, the bad, and the ugly – about using ChatGPT in writing your applications.
In this episode, you’re going to hear from:
- Clare Norton, Columbia Business School Senior Associate Dean for Enrollment Management
- Shari Hubert, Associate Dean of Admissions at Duke University’s Fuqua School of Business
- Blair Mannix, Executive Director of Graduate Admissions at the University of Pennsylvania’s Wharton School
- Eric Askins, Executive Director of Full-Time MB admissions at UC Berkeley Haas
- Dean Robert Salomon, the inaugural Dean at Stern at NYU Abu Dhabi.
I’ve asked this question of almost every admission director I’ve spoken to, so these responses represent a sample. And there is some difference, there’s more differences on this one than on some other questions, but in any case, I think you’ll see that there’s cautious acceptance of ChatGPT, with several caveats and warnings for applicants.
Clare Norton, Columbia Business School’s Senior Associate Dean for Enrollment Management. [2:43]
[CN] I think ChatGPT is a tool, and there are many, many tools that we have now that we did not once have. At some point in time we thought to ourselves, “If people use a calculator, will they understand math?” Yes, they do still understand math, and in fact, probably, probably higher level math than they understood before that was utilized broadly. And I suspect ChatGPT will be quite similar. We’ve made it very clear to students in our application process that it’s a tool that can be utilized, but generative AI is not something that can write the whole answer. It’s the kind of thing that could do some editing for you or provide you with some ideas to make sure that you’ve touched upon, but that the work must be your own. So from that perspective, I think we’re quite clear.
But also I think, actually back to what we were just talking about, the best applications are reflective, truly, of the individual.
And our essay questions in particular. We are really asking you to say, “For you personally, what is it about this that is going to connect, assist you, help your growth, engage you in new ways?” And generative AI is not capable of saying that in a way that is authentic.
So that’s really what we’ll be looking for. And so I think, again, it’s nice to have something, maybe check your grammar, right?
It’s there for those kinds of things, right? But it’s never going to give an answer that can tie together across the application.
I have a daughter who loves to bake. Packaged cookies just don’t taste as good as her fresh made cookies out of the oven or bread or whatever it is that she likes to bake. [4:29]
[CN] Yeah. There’s a little bit of her in that recipe that just can’t be replicated.
Same for this. So yeah, I think it’ll actually be exciting to see what it does. And we as a school are obviously, as I mentioned earlier, thinking about what kind of training do our students need and how do we engage with it as a tool and make sure that they know how to engage with it as a tool and think about what kind of management is required and what data is missing from the data sets that these kinds of tools are drawing upon? So there are lots of great and important questions for us to engage inin the classroom, so we’re certainly not shying away from it. It’s something we think is important. But I think it won’t be problematic from an admissions perspective, given again, the fact that A, we’re letting people know what our honor code is and how we expect them to utilize it, and that we also really are looking for personal insights in our essays
Shari Hubert, Associate Dean of Admissions at Duke University’s Fuqua School of Business. [5:36]
[SH] First of all, let me just say I can only speak to our policy within admissions as it may vary across the university and then the Fuqua School in terms of classroom use. It’s really going to be up to the faculty to decide that. But within admissions, allowing the use of AI in their application, and we have decided to allow it. It felt like the way to be the most inclusive while still requiring that applicants authentically represent themselves. We see a difference between plagiarism and the use of AI, in that plagiarism is explicitly using material created by someone else. While we expect that the use of AI, at least in terms of how they might use it to answer our essay questions which are unique to Fuqua, the use of AI, it has to begin anyway with this level of personal reflection. I mean, to answer our essay questions, you need some level of personal reflection. You need your own kind of content and your own lived experiences to inform it.
But we know that AI could be useful in terms of helping people organize their thoughts or represent them better, differently, through the use of AI tools. We see, similar to how people use Grammarly, or they may have friends who are English majors and they ask them to review their essays, or they may use admissions consultants to say, “Hey, take a look,” provide some coaching and guidance around their essays. So again, we view this as a tool that enhances the process, but should not and does not replace the requirement for authenticity, and the use of your own material.
And so in our minds, and I like to say, “AI at Fuqua stands for authentic individuality.”
And we’re going to assume positive intent and that applicants are ethical and they’re good agents in this process. And so we do require that your application be a true and accurate reflection and representation of your lived experience and exclusively your own. And then we do, like you said, use plagiarism tools. So for us, all essays are scanned using plagiarism detection software, but again, we see a difference between plagiarism and the use of AI tools. So we have a long kind of disclaimer about how expressing your ideas by using verbiage that’s not sourced right, is improperly credited, is a violation of our honor code, and it is grounds for denying application.
Blair Mannix, Executive Director of Graduate Admissions at the University of Pennsylvania’s Wharton School. [8:03]
[BM] Faculty have done great work on the use of ChatGPT in passing Wharton exams. There’s a couple of fun articles about that too. But this is where the office of MBA’s admission stands, this is where I stand. My view, personally, is being fearful of ChatGPT and what it means for society is like being scared of email. It’s coming, it’s here, it’s an efficiency tool. People will be using it. I don’t believe it’s the right move to gatekeep using an efficiency tool, especially if we expound, if we say that we’re cutting edge business leaders. Cutting edge business leaders outside of our walls should be using ChatGPT to speed up their work. Now ChatGPT, as we all know, has limitations. There’s studies in law schools that they put fake legal precedents on some of the answers they do.
You have to check it. But I think it’s only going to get better and so I think it’s an efficiency tool. Behind the scenes we use it for our own work, faculty at Wharton use it for their own work. I think people should use it going forward. And so I have no reason to or no plans to put a disclaimer on our application saying, “Please don’t use ChatGPT,” because we all know it’s not good. You’re going to have to work with it. And I think that’s okay.
Eric Askins, executive director of full-time MBA admissions at UC Berkeley Haas. [9:26]
[EA] I think I’ve gone on the journey that a lot of my peers have gone on. And the first piece of this journey was, well, I hope our fraud software can catch it. And I think a lot of the world has gone on this journey with us where you stop and then you say, “Well, this is a tool. This is a tool like the calculator is a tool.” That’s probably the common thread I’ve heard. I’ve already had typing in an email and I’m getting suggested next three sentences. This is where we are. The tool exists.
Now I am still going to suggest that there’s no tool that’s going to tell me your career goals. Now, that tool might help you articulate those career goals a little bit better, but those goals still have to be yours. There’s no tool that’s going to tell me the moment that made you feel alive or why it gave meaning to you. It may be that that tool helps you frame your thoughts, put those pieces together in a way that’s cohesive. If English is not your first language, and you’re trying to organize your thoughts in a way that would give you the tools to succeed, it could very well be that this is a really useful tool to organize. But those core thoughts have to be yours. I think that’s key here, and I don’t think that we’re going to move on that concept. But those core thoughts and ideas have to be yours.
And then if you are going to use the tool, I hope that you use it well. I hope maybe the thing you’re demonstrating to me is your expertise in the use of the tool, because I will, and we have seen already, poorly framed and poorly worded things that don’t really seem to capture the individual.
This is probably the first year that we’re starting to see that.
That makes sense that this is the first year you’re seeing it. I’ve played with it a little bit. And I’ve said this before on the podcast, if you use it blindly, you’re going to produce drivel that’s very generic and not very meaningful. If you use it either to edit your work, perhaps to generate some ideas or to help you structure an essay, but the ideas are your own, perhaps it has value, but you’re still going to spend a significant amount of time on it. [11:08]
[EA] I’m one of those folks that believes the magic happens in the editing. I know other people think it happens in the writing, the idea generation. I think it’s the moment where you come back and say, “Oh, well, now I see how those pieces should fit together.” And so with that in mind, I understand that the tool may be used.
We have a statement at the bottom of our application. We haven’t changed it. We’ve had it for a while. It says, “The work product seen here is mine and mine alone.” I think folks should be able to answer that honestly.
The work product here is mine and mine alone. Now if that means that they used the tool to take their ideas and put it on paper, and then they reorganized it to reflect the story that they wanted to tell, and they feel that that is theirs, they were the producer of the ideas. They were the producer of the finished product. They used an intermediary tool the same way you might use a spellchecker or a grammar checker. I’m going to have to just accept that that’s the world that we’re in today. I don’t think there’s any magical tool that solves that one yet. Generative AI is probably the best tool to catch generative AI.
But I am going to focus on the content. And as long as the content’s strong, I think that that’s going to be in the candidate’s best interest.
Dean Robert Salomon, the inaugural Dean of Stern at NYU Abu Dhabi. [12:49]
[RS] As a teacher, it bothers me a little bit that students might rely on ChatGPT and not hand in their own original work, that they use it as a crutch, especially those people who use it when they’re short for time, they’re being a little lazy. They were like, “Oh, well, I’ll just have this program do it.” So that bothers me a little bit. But I think now the onus, again, as an educator, the onus then is on us to create assignments that maybe leverage the benefits of ChatGPT and bring it as a tool to help enhance learning.
So we’re all still trying to figure it out together. In the meantime though, we can’t have students just using it to copy and paste to their assignments. And we have a policy against that. On the applicant side, we do have a policy. One of the things that we ask our students is to verify that they have not received any outside support in preparing their essays in their application and that students who are found to have gamed the system and to have used outside to support, we can revoke their admissions.
So we do have those policies, but again, the onus is on us to create prompts that make it difficult to use things like ChatGPT. On the admission side, we want to know who the students are and we don’t want to know what ChatGPT thinks. We want to know what you think. And in the classroom, we want you to learn. We want you to push yourself. We want you to enhance your capabilities. And you can only do that if you really are putting in the effort and not relying on an outside tool to do it for you.
Right. I think there’s also a difference between using an outside tool and relying on an outside tool. [14:42]
[RS] Right. Yeah, I don’t mind them using outside tools. And I’ve been designing a little bit assignments that leverage ChatGPT in a way to help students learn.
Would you be willing to share an example? [15:05]
[RS] Yeah. One example could be you actually put the prompt in ChatGPT that you want students to answer and then you ask them to critique the response from ChatGPT.
What did ChatGPT get right, what did ChatGPT get wrong, and why? Another thing that I do in my classes is I have students work together in class and I go from group to group, we have discussions in small groups. And they have to think on their feet. So they’re not prompts. The discussions that we have are not prompts that ChatGPT would know how to answer. So those are the ways that I approach it. But yeah, I mean, you’re exactly right. We want students to use it as a tool. We don’t want students to rely on it to do their work for them.
The consensus is…[15:54]
If students can use ChatGPT, so can profs and the administration. After all, what’s good for the goose is good for the gander. To summarize, most of the MBA admissions committee members viewed ChatGPT as a tool, Some with more trepidation than others. But in any case, like all tools, it can be used well, and it can be used very poorly. Relying on it blindly is a recipe for generic, vapid, empty, boring essays, probably a rejection. Providing ChatGPT with a lot of input and specifics coming from you can help. However, it may take as long as you just sitting down and writing your essays.
If you choose to use ChatGPT, first, really think about what you want the schools to know about you. ChatGPT can’t discern what you’re most proud of. It can’t know, unless you tell it with a fair amount of detail, when you assumed leadership responsibility or what are the life challenges that you have overcome? It can’t say, without you giving it lots of information, where you excelled or when you handled an interpersonal challenge with finesse. Recognize its limitations and first put in the necessary thought so that you can have an application that is thoughtful, authentic, and effective.
If you’d like help in presenting the best of the authentic you, please contact Accepted for guidance of presenting your best self and polishing that gem of an application. Discover how Accepted’s experts can help you and take advantage of an initial free consultation.
Related Resources:
- How to Get an MBA at Columbia Business School, podcast Episode 528
- How to get into Duke Fuqua, podcast Episode 536
- What’s New at Penn’s The Wharton School. And How to Get In, podcast Episode 545
- How to Get Accepted to UC Berkeley Haas Full-time MBA Program, podcast Episode 547
- Stern at NYU Abu Dhabi: A Full-Time MBA in the Middle East, podcast Episode 549
Subscribe: