How Generative AI is Transforming Academic Work
Generative AI is changing the way academics work and learn. Is it cheating or a cognitive revolution?
Generative AI is a powerful technology that is transforming society. It will likely change the way we all work. In a recent survey of over 80 students and professors at TU Delft, more than a quarter already use generative AI —and more than half plan to do so in the future.
By next year, once Microsoft completes the integration of OpenAI image and text generators within Word and PowerPoint , generative AI will be completely pervasive. How will this affect academics?
AI Cheating
I just read about 40 student papers and they were suspiciously much better than last year. Indeed, generative AI is already widely used to, ahem, scam teachers everywhere. Some see a dystopian future where students automatically generate essays with AI —and teachers automatically grade it with AI. (Wait, maybe that’s a utopian future: don’t we want a future where the robots do all the hard work?)
Here’s the issue: Generative AI poses a genuine problem when it comes to accurately assessing a student's knowledge. If a student uses generative AI to automatically complete writing assignments, they may be misrepresenting the knowledge that they actually have. This threatens the integrity of many academic systems which rely on the ability to measure learning outcomes.
So, don’t be surprised if we return to the widespread use of handwritten essays in large examination halls. Observing students writing on paper may be the only way to ensure that the student submits AI-Free writing. (Students better brush up on their penmanship...!)
AI Writing as a Reasoning Engine
Let’s be clear: writing academic work with AI is not necessarily cheating. I think of ChatGPT as a "reasoning engine" or sparring partner that can assist academics in analysis, brainstorming, ideation, and problem solving. Yes, it can also help with writing clearly—believe it or not, writing clearly is a real challenge for many scientists. In short, chatGPT can help scientists reason through complex concepts and help them communicate more effectively.
The caveat, of course, is that chatGPT can be disastrously and confidently wrong. Skepticism and critical thinking must be exercised continuously. Additionally, it is important to know that there is a real learning curve when using generative AI: knowledge and experience can greatly improve the outcomes you get. So keep playing with it! We are only beginning to understand how useful Generative AI can be in different domains.
Forming Local Communities of Practice
As generative AI continues to evolve and transform academic work, I am working with the ErasmusX organization to help build communities of academics to share knowledge. We want to create venues for educators, students, and technologists to come together and explore how to effectively use generative AI and to make sure it is aligned with our values. Generative AI poses serious ethical implications and it is important to develop best practices around its use. In upcoming weeks, I’ll be sharing an open invitation to academic AI writing workshops.
Join the Conversation
I invite you to join the conversation about the impact of generative AI. What are your thoughts on how AI is transforming academic work? How do you envision using generative AI in your own work and learning?
Thanks for reading, and I look forward to hearing your insights!
Best,
Derek
Derek Lomas is an Assistant Professor of Human-Centered Design at Delft University of Technology, where he leads the "AI and Experience Design" group in the faculty of Industrial Design Engineering. As a designer, futurist, and cognitive scientist, he is passionate about aligning emerging technologies with humanity. His current work investigates the potential of generative AI to transform the way we work, learn, and play. Through his research, he seeks to better understand how to design technologies that can enhance our humanity rather than replace it.