C-Level View | Feature
Preparing for the Classrooms and Workplaces of the Future: Generative AI in edX
A Q&A with Anant Agarwal
As the chief platform officer of 2U, founder of edX, and an MIT professor, Anant Agarwal has been at the heart of technology change for most of his career. Here, CT asks for his perspectives on the impacts of generative AI in higher education environments and what edX is doing to help lead the way in the productive use of innovative new tools based on the technology.
Mary Grush: Despite some initial hesitation about the technology theoretically usurping humans in the teaching and learning context, do you now see more acceptance and trends that seem to indicate that generative Al can be profoundly good for higher education in the long term?
Anant Agarwal: Yes! Just as the calculator did not displace the learning of mathematics, but rather became a tool that students use to enhance their learning, so will generative AI become a tool for learners and educators. I think of it as a helpful avatar or copilot, and not a replacement for the fundamentals of teaching and learning.
For example, learners will be able to work with and alongside generative AI to help bolster and build their writing skills. But will AI entirely replace the higher-level writing skills, such as applying and making arguments with reason or persuasion? No. Those are uniquely human skills. Generative AI can serve as a tutor or helper, providing prompts, hints, or recommendation summaries — as does Xpert, an AI-powered learning assistant embedded in the edX platform. Learners can engage generative AI to augment the learning experience in a way that is additive to them, by quizzing themselves or reviewing key concepts. Again, I see generative AI as a valuable and accessible tool that doesn't replace the human element in learning or in teaching.
For educators, generative AI can be an incredible copilot. It can help grade essays, create new content, build new and more diverse problem sets, and even brainstorm creative ways to engage students. All of this makes the learning experience more personalized. But will generative AI replace the richness and dimensions of a human instructor sharing knowledge and teaching with skills imbued with his or her own unique personal perspectives and passion? No. Those are still uniquely human skills.
The underlying concept here is help: Generative AI can help educators as an additive tool, but it won't replace them. As this generation of learners embraces generative AI in their learning experiences and in their lives more broadly, educators will need to follow suit.
Generative AI can help educators as an additive tool, but it won't replace them.
Grush: How would you address the more foreboding and probably very real concerns you've heard educators express about Al, specifically in higher education environments? How can those in leadership positions focus on, and plan on, the shift to Al in ways that ensure the best of change?
Agarwal: As I discuss AI with colleagues, partners, and learners, I'm reminded a lot of 2012 — aka "the year of the MOOC" — when many people were concerned that digital learning would totally upend traditional in-person learning. At that time, those of us developing MOOCs were especially thoughtful about how we provided digital learning. At edX, we open sourced our technology as Open edX so that everyone could access it. We also worked on creating blended models of learning that enabled learners and faculty on campus to benefit from digital learning as well.
In a similar manner, we want to consider AI carefully, and utilize it responsibly and transparently.
Just as I was, in my time as director of CSAIL, MIT's Computer Science and Artificial Intelligence Laboratory, I now am deeply optimistic and encouraging of higher education institutions to embrace the clear potential of generative AI in developing innovations in teaching and learning. New technologies can be scary, particularly ones that are developing as quickly as generative AI. But universities and companies can and should be adopting principles and guardrails so they help faculty, students, and administrators alike get on the same page for how to use these tools well in their day-to-day work and learning.
As leaders of institutions, it is important to get on board early to experiment with AI, because you can then help shape the future responsibly, rather than having it done for you by others who may not necessarily have your best interests front and center.
It is important to get on board early to experiment with AI, because you can then help shape the future responsibly, rather than having it done for you.
Grush: In light of recent announcements, can you tell me a bit about the plans and expectations for generative Al in edX? What do you think will be the best outcomes of generative Al in edX?
Agarwal: The best outcomes when leveraging generative AI can be put quite simply: to build cutting-edge capabilities that provide a more personalized experience for learners that will be highly impactful in maximizing learning. The latest announcements we've made at edX around AI are grounded in this belief.
With the newly-available edX plugin for ChatGPT, we're empowering learners to access in-depth learning resources on select advanced topics — beginning with Circuits and Electronics. As learners engage, the plugin will instantaneously provide course recommendations, content, videos, and quizzes. Similarly, Xpert, the AI powered learning assistant on the edX platform, was designed to provide learners with real-time, personalized academic tutoring and customer support as they engage in online coursework.
We believe these latest developments build on edX's leadership in delivering quality student support and driving impactful outcomes. There's much more to come, and we'll continue to seek out opportunities to help learners everywhere to access the education and support they need to fuel their ambition and achieve their goals.
Grush: Can the work on Al in edX be a good model for others with responsibility for IT programs in higher education?
Agarwal: Absolutely, that is our goal! We are committed to creating a dialogue with our partners and the larger network of edtech innovators around what it means to integrate AI into education responsibly. Like any technological advancement, we have to be thoughtful about how we implement AI in the way we work. We also want to ensure that we provide learners with the opportunity to learn about the technology, too — that's why we are so excited to offer a range of programs for learners from our partners on edX. That includes everything from a fully online master's degree in AI for $10,000 from a top U.S. public institution, to free courses covering "Prompt Engineering and Advanced ChatGPT." Our goal is to be sure there is something for everyone.
We are also working with our institutional partners on responsible and transparent use of AI by creating a dialogue on the best ways of using AI. The core principles that we develop will be useful for AI adoption by many other institutions.
The biggest step we can take as an industry is leaning into the possibilities and helping as many teachers and learners as we can prepare for the myriad challenges and opportunities ahead. Generative AI is a powerful, transformative technology, and those who learn how to harness that power will be the best prepared for the classrooms and workplaces of the future.
