Responses have been edited for length and clarity.
Professor Hatim Rahman
Kellogg Assistant Professor of Management & Organizations who researches how AI impacts employment relationships
ChatGPT, like any other technology, serves as a mirror to our society because it reflects what we value and prioritize. When I look into this mirror, I see an amazing technical achievement, but I also see an appropriation of knowledge, labor and natural resources, hype and unequal access. A major source of ChatGPT’s training data is found on the internet, which includes copyrighted materials and data that many people and organizations did not give consent to use for training an AI system. When ChatGPT provides you an amazing answer, it is likely appropriating and profiting from someone else’s knowledge and labor without giving them credit or compensation. A recent analysis found that a copyright symbol was found 200 million times in a dataset that many AI models are trained on. ChatGPT’s servers also consume precious natural resources every time we use it; this is a factor we cannot overlook when trying to address climate change.
As much as it is easy to believe ChatGPT will replace many jobs, history suggests otherwise. Since the 1950s only one job has been eliminated by automation — the elevator operator. The most likely outcome is that ChatGPT’s impact on jobs will be shaped by complex, often unanticipated, factors. Many people may change the way they work (not always in ways that help), but it is unlikely to lead to mass unemployment. Further, studies show, despite their technical prowess, most new technologies — including the internet, personal computers and smartphones — have had a limited long-term impact on productivity.
An uncomfortable reality for many technologies in the past few decades is they often concentrate wealth and opportunity among the elite in society, further exacerbating inequality. Thus far, there is nothing to suggest ChatGPT will be different. Already, only people who can afford to pay a monthly fee get access to its updated, more advanced model.
Professor Randy Hlavac
Medill IMC Professor who teaches Digital, Social and Mobile Marketing
Content strategy and digital marketing are vital components of today’s business landscape. With the increasing amount of content available online, it’s becoming more challenging to stand out and make an impact. Artificial intelligence, specifically ChatGPT, is one tool that can revolutionize content strategy by providing content strategists with a new way to generate insights, define specific audiences and create content using a multimedia approach.
While Northwestern students know of ChatGPT, they don’t understand the creative power they can achieve by training ChatGPT effectively. If used correctly, it can develop anything from strategic and competitive insights to forms of content including articles, video scripts, blog posts and “outside-the-box” concepts, social media posts, Tweets, hashtags, emojis and even entire websites.
Using ChatGPT in content strategy can significantly accelerate a marketer’s ability to create engaging and effective content. ChatGPT can generate innovative insights and ideas, help define specific audiences and aid in creating all types of content using a multimedia approach. By utilizing ChatGPT, students can create high-quality, engaging content that resonates with their audience and generates positive business outcomes.
ChatGPT helps marketers move from strategy to concept to content to marketing — all the steps required to create exciting, engaging content and to make a target audience aware of it. Marketing at the speed of AI!
Emma Estberg
McCormick fourth-year, Applied Mathematics
In one of my courses this quarter [Data Engineering Studio], my syllabus included a sentence about ChatGPT along with the usual statement programming classes include about sharing code and using the internet. To my surprise, we were encouraged to use ChatGPT. I didn’t think much of it until the first class when we were going over the syllabus and my professor [Diego Klabjan] repeated the sentiment. He began gushing over the usefulness of the tool while simultaneously warning us the code it produces will likely include errors.
The course uses a flipped classroom environment, and I was shocked to hear a text-to-speech voice when I pressed play on the first video. The next class, my professor explained — he had made the slides, then used ChatGPT to generate a transcript, then fed it to a text-to-speech to use for the video. During the demo that period, he pulled up ChatGPT to ask it to write some code. In another class, he asked it about coding syntax. He chooses to use ChatGPT over and over, so much so that now there are snickers in class any time it is mentioned. And each time, he points out the errors in the results.
What concerns me most about ChatGPT are its errors. I can imagine people will use it and take whatever the AI spits out to be the truth. People do not fully understand the process behind the generative pre-trained transformer and how it uses prediction to determine which words should come next. I keep coming to the idea that we are not ready — the general public isn’t ready — for this jump in technological advancement.
Hannah Cole
Medill third-year, Journalism
The usefulness of ChatGPT depends on how people use the platform. Using ChatGPT to create recipes from the food in my fridge is helpful, and talking with fictional characters when I’m bored is incredibly entertaining. I have heard mainly from computer science majors that ChatGPT helps create or check code. However, if you’re a humanities student, ChatGPT should not be your friend. After reading numerous discussion posts and peer-editing essays that use ChatGPT, it’s obvious. Small grammar mistakes and accidental word swaps are everywhere. Not to mention, ChatGPT doesn’t consider your assigned course readings.
That said, every one of my humanities courses changed its syllabi to prohibit ChatGPT. Meanwhile, computer science courses are encouraging students to use it. My professors seem incredibly worried about academic integrity, yet ChatGPT should be the easiest to catch for essay submissions. The fervor and worry surrounding academic integrity undermine students’ ability to think for themselves and use their judgment not to use electronic essay writing websites. These platforms have been around for ages, and ChatGPT isn’t bringing anything different to the table in that respect. Perhaps for computer science majors, it’s life-changing, but I wouldn’t know about that. Most students at this school know that essay-writing platforms don’t produce anything good or usable, so it’s unclear why some professors are so worried about that changing. Perhaps in the future, the platform will figure out a way to factor in readings and produce a perfect five-page, double-spaced, analytical essay. However, until that becomes a reality, the anxiety over ChatGPT is too high.
ChatGPT's Response
Prompt: Write a 200-word essay about the benefits or dangers of using ChatGPT
ChatGPT is an innovative technology that utilizes artificial intelligence to interact with humans. While there are several advantages of ChatGPT, there are also some potential risks that need to be considered.
One of the primary benefits of using ChatGPT is that it can provide fast and accurate responses to a wide range of questions. Since ChatGPT is trained on vast amounts of data, it has the ability to comprehend and respond to various queries, making it a powerful tool for customer service and research.
Furthermore, ChatGPT has the potential to automate many tasks, saving time and resources for businesses and organizations. By using ChatGPT to handle routine tasks such as scheduling appointments or answering frequently asked questions, companies can free up their employees to focus on more complex and high-value tasks, ultimately improving overall productivity.
However, one potential danger of using ChatGPT is that it can be utilized to spread misinformation or propaganda. Since ChatGPT can be trained on any dataset, it is possible for bad actors to use it to promote false or harmful information.
Another potential danger is that ChatGPT could be used to replace human workers, leading to job loss and economic inequality. While automation can be useful, it is crucial to ensure that it does not result in negative consequences for society.
Overall, while ChatGPT has several benefits, it is essential to use this technology responsibly and ethically to ensure that its advantages are realized while minimizing its potential risks.