341: How AI is Changing Education: The Future of Accessible Learning with Marni Baker Stein, Coursera CCO
31:32 Share

And I think AI is going to help us do that. It's going to allow us to meaningfully and powerfully personalize and contextualize and make much more dynamic every learning experience, as well as to more intentionally link learning experiences and credentials to what's going on in the world.

Good morning, good afternoon, or good evening, depending on where you're listening. Welcome back to AI and the Future of Work. I'm your host, Dan Turchin, CEO of PeopleRain, the AI platform for IT and HR employee service. As you know, we bring you fascinating conversations with the leaders in this community, the AI community, investing in better employee experiences. Every Monday, new episodes drop. Hopefully, you know by now, we publish a weekly newsletter every

If you're not already subscribed, you should be. There's a link in the show notes. Go join us there. We share additional fun facts, tips, some things that don't always get shared in the weekly episode. If you give us a like and a rating on Spotify, Apple Podcasts, or wherever you listen, we greatly appreciate it. If you leave me a comment, I just may share it in an upcoming episode.

Like this one from Manpreet in Bangalore, India, who's a software developer at a tech services company. Manpreet's favorite episode is the one from season three about building a unicorn startup with the amazing and humble CEO of Zapier, Wade Foster. Go check that one out if you haven't heard it already. We learn from AI thought leaders weekly on this show. The added bonus, you get one AI fun fact each week. Here's today's fun fact.

Mahmoud Azar writes in the daily Saba online about AI and the future of education. In it, he says AI is transforming education by offering tools and raising serious risks, making ethics-based AI literacy essential. Teachers can rapidly generate various types of content related to their lessons and use them in classroom settings. In this new landscape, the workload for teachers is undergoing a fundamental shift while conversational tasks are decreasing significantly.

They must focus more on the individual development of each student. There's the potential to transform the educational environment from a traditional setting into a much more innovative and interactive learning space. AI-driven tools have the potential to reduce one of the most significant challenges for teachers, the disparity in students' academic achievement. In this context, personalized learning emerges as a possible solution.

AI systems bring not only benefits but also risks beyond concerns about data privacy. The most significant risk is that the content generated by these systems may or may not be accurate.

Of course, we'll share a link to that full article in today's show notes. My commentary, AI is part of the new fabric of society because we rely on the education system to train students to make positive contributions as citizens of the world. It's where we especially need thoughtful conversations about the evolving relationship between humans and machines, students and teachers.

Expect more experiential learning, more assessments that focus on critical thinking and communication, and more Socratic dialogues between students and teachers. All of that is aspirational today, but I believe we'll look back on this decade as the one where human ingenuity advanced more than it has in the past thousand years. And that is very germane to the conversation we're going to have today with Marnie Baker-Stein, who's the Chief Content Officer at

Coursera, where she oversees the company's content and credential strategy and partner relationships. Coursera, of course, is the online learning platform, one of the pioneers in the space. Started in 2012 by Daphne Kohler and Andrew Ng from Stanford, it's used by more than 175 million learners across more than 6,200 campuses, businesses, and governments to learn skills and earn degrees.

Marnie has more than 25 years of experience producing and scaling online and hybrid education programs. Prior to joining Coursera, she was chief academic officer and provost at Western Governors University, where she led its four colleges, serving more than 135,000 students. Before that, Marnie held leadership positions focused on access, student success, and program design at places like the University of Texas, Hookham Horns, Columbia University, and the University of Pennsylvania.

Marnie earned her PhD in educational leadership from the University of Pennsylvania. Go Quakers! And without further ado, Marnie, it's my pleasure to welcome you to the podcast. Let's get started by having you share a bit more about that illustrious background and how you got into the space. Sure. So thanks. Thanks for having me. I'm super happy to be here. I have been

weirdly working in online hybrid and sort of digitally enhanced learning since the 90s, since the launch of the first mosaic website, and have had a really amazing opportunity to work across Ivy League institutions, a big state system like the University of Texas, and a totally online competency-based university like Western Governors University on figuring out how to do that

to do it well, to do it in a learner-centered way, and to do it in a way that is not just an okay alternative to an on-ground experience, but in ways that truly compete and in many ways advance learning beyond what can happen in that face-to-face environment. Now, in the intro, I talked a little bit about the complicated future

of education. From your perspective, in one of the places that really, as I mentioned, has been a pioneer for the last 14 years, 15 years, what do you see as the future of education? Well, I truly agree with what you said at start that AI is the engine of the future of learning. It's transforming and it's going to continue to transform learning from static, one-size-fits-all

learning experiences, which is really all that we could do in the past, whether we were online or whether you were in a brick and mortar campus, like we just didn't have the ability to

personalize and contextualize learning at scale. And I think AI is going to help us do that. It's going to allow us to meaningfully and powerfully personalize and contextualize and make much more dynamic every learning experience, as well as to more intentionally link learning experiences and credentials to what's going on in the world, what's going on in the job market in ways that just were...

You know, we couldn't even fathom even just two, three years ago.

So let's say for the sake of conversation, there are three key types of constituents. When we talk about the future of learning, there's the learner, obviously, there's the educator. And then let's say there's the institution who's providing the content or the framework. I would go out on a limb and say every human is a stakeholder in one of those three categories. Maybe if you could take each one, if that would be okay, and talk about what we should expect to

to see change starting with maybe the learner? Well, I think, I guess I would just back up and say that I think what we'll see change beyond just AI powering new learning environments is we are going to see micro-credentials, shorter, stackable, flexible learning experiences as the new currency.

of education. I believe that will happen for lots of different reasons. And I think that will be actually a really powerful change into the future. So how does that impact learners? I think for learners, what it means is they're going to be a lot more points of entry for them to learn and master skills and competencies and mindsets that

and be introduced to new applications for these skills and competencies and mindsets than ever before. I really believe that education, because of the scale constraints, has always really targeted one learner profile. I think there's going to be room for a lot more learner profiles into the future. I'm truly excited about that.

I think for instructors, teachers, professors, it's going to mean that things that they started to experiment with,

whether it was pre-COVID or during COVID around online learning or digitally enhanced learning or hybrid learning, that they're just going to have like X more powerful tools to do creative stuff as they're putting together learning experiences for this more diverse learner audience. And they're not only going to have more interesting tools, but those tools are going to make their work not only more creative, but I think more efficient as well.

And for institutions, institutions are going to have to think about, well, like, what does this mean, like, for our modalities, right? So whether, like, they're going to be considering, should we have more online programs because of this ageism?

AI-powered, more personalized, stackable learning experiences? Are we going to have more hybrid components? Are we going to use our physical infrastructure differently because of these tools? Could we kind of turn upside down what a campus is into the future? So I think they're going to have a lot of interesting questions to ask and answer themselves for the same reason. And I think you can take a kind of

very positive, optimistic lens on this and say, that's cool stuff all the way across. It's going to provide opportunities for evolution and transformation in education for learners and teachers and institutions. And you could also be very pessimistic as certain people are and say, you know, like, well, change is bad and we're going to lose a lot when this happens. And, you know, I think we're seeing both right now. We're seeing people who are...

learners, all stakeholders, learners, teachers, universities who are super excited and getting really creative about how they can use these tools to do things that are going to change the world and change everything we know about teaching and learning. And then we have other people who are hanging on to the past. And I think that's probably

a good balance, right? That's a good balance that protects us from moving too fast and creates affordances and constraints for the creativity that I think will then be pressure tested in important ways as we move forward. From the perspective of the institution, should they feel institutions feel threatened or is the increasing presence of AI more of a threat or an opportunity for the business model that is higher ed? So a lot of people think

Ask me this question. And it usually starts with something like Barney, you know, is the four year degree dead or something very controversial like this? And my answer is always no credentials and these institutions themselves are

are not dead. They're probably, they're more relevant than ever. And their mission, from my perspective, couldn't be clearer. But I do think that mission is being redefined. Their learner populations are changing. The tool sets that they have to contemplate what the future of education will be are changing. And they're going to have to grapple with

looking different in the future than they have for the last 50, 100, 300 years and really moving into a kind of next state where they are more modular, where they are more affordable, where they are tightly aligned to sort of the new skill sets of the future, both technical and non-technical, and where they understand that in the future, people are going to have to learn lifelong, not just

have coming of age experiences between the time they're 18 and 24, but really could be coming back again and again to that university experience, whether it's online or on ground and taking advantage of

of what universities have to offer and that community that they provide as well. So I don't think they're dead, but I do think they are headed toward a massive transformation that is much needed. And we need leaders in higher education who are going to find that exciting and go with that flow.

In the HR community, there's a lot of talk these days about a future of work that involves a focus on skills versus jobs. The extent that resonates with you or, you know,

how you're exposed to the learning community. What does that mean? And is that an accurate way to frame the future of education? Yes, I hear the same thing, especially from our enterprise partners in Coursera for Business, that the focus is on skills. And I, of course, I come from a background of skills-based and competency-based education. So I do believe that that is true, or at least partly true. At Coursera, what we're really trying to do is

understand how new technologies and the evolution of technologies, AI in particular, we are interested in how those technologies are impacting occupational bands, occupational group bands. And

Our focus is on skills, but it's skills contextualized to different sectors and different occupational bands. So that's kind of how we're looking at it at Coursera. We're seeing across those bands, Gen AI skills exploding. We've seen 12 enrollments in AI courses across sectors every minute in 2025.

We have over 700 Gen AI courses live, and so we're constantly creating them both for general purposes, for developer purposes, but also cross skill applications as different companies in different sectors are beginning to deploy Gen AI in really interesting ways. We are seeing a real hunger, a

across these different occupational bands for contextualized skills that are grounded in not just awareness about the skills, but very specific occupational related applications. So everybody wants to learn OpenAI or Lama or

whatever the tool is, Canva, whatever the tool is that is really critical, but they really want those occupationally grounded applications. And the other thing we're seeing is some of the areas where we're having our fastest skills growth are these human enduring skills. So what I think is really interesting is, yes, skills are important. People want practical occupationally grounded applications of those skills, but

And as this transformation is happening, I think a lot of people are also thinking, hey, you know, these tools are not that hard to use. I better brush up on some of my human skills and sort of my enduring understanding of the world around me and how to do research and how to ask good questions and how to

Think in interesting quantitative ways around the world around me. So we're seeing a lot of popularity of those courses as well. It's not just the tech courses. So that's the way I see it. Skills are really important, but skills without context are almost meaningless. You really need that context. And what a lot of people are hungry for is the context that's associated with the job they have.

or the job they wanted, the promotion or the job that they want to get. So it's a much more satisfying answer. And I usually get to a similar question. Okay, so I got to get your take on this. So we had a great recent conversation with Dave Marchick, who's the Dean of the Kogod School at Business School at American University.

in DC. And he was talking about kind of renovating the traditional curriculum, at least for undergrad business students, renovating it to account for the fact that a lot of the skills that employers are expecting in the modern workplace are going to involve an understanding of some of these AI-first tools. We also had a conversation recently

with a gentleman named Chris Caron, who's the CEO of a company called Turnitin, which is ubiquitous in higher ed. I think over 50,000 institutions use it to detect plagiarism. Is AI an impediment or a threat or a way to cheat for students? Or maybe more like Dave Marchik says, it's part of the new fabric of society, and we owe it to society to train these skills and not

you know, not have students perceive it as a, you know, something that they're trying to hide their use of. Yeah. Yeah. I mean, well, AI is certainly not the enemy and it's, it's not an impediment. I truly believe it's the new co-pilot for learning and teaching. And, you know, can it support bad action? Can it support cheating? Absolutely. Cheaters have always been out there and cheaters are going to cheat. But,

But we are seeing in our own research that faculty are overestimating cheating. So a lot of faculty are feeling like cheating is going on. But really, in the actual case, that is not perhaps as common a

an activity as they believe it is. Still, we need safeguards and we need to make sure that learners are using these tools to effectively learn, to contextualize their learning, to make their learning meaningful to them. And as I said earlier, that we're inviting more learners into the fold to learn things that help that personalization and sort of

of like just-in-time AI-powered tutoring can really help. So is there cheating? Yes. Is it overestimated? I think yes. Does that mean we don't need safeguards? Absolutely not. And at Coursera, we're investing in AI tools that really support integrity. So we have dynamic assessments, AI tutors, proctoring that's powered by AI that really help us to understand not only are you the person taking that assessment, but

but also did you really learn? Were you the one who really learned the topic that was being assessed? We're also teaching educators how to use AI responsibly. We have a new course by Wharton and OpenAI that's on this topic. So we take it really seriously. We think AI is a problem, perhaps on the cheating front, but it's also the answer. And we certainly can't use AI

The fact that some learners may be cheating now as a reason to not move forward with AI-powered innovation in the learning space, that's really going to democratize and drive access for learners more broadly than we've ever seen before. Everyone listening, replay that answer. That was a really good answer. So I believe it's an opportunity to introduce innovation.

ethics into the curriculum and make, you know, really celebrate this opportunity that we have to not assume that it's going to increase or even decrease cheating, but let's have the dialogue about what cheating is. And because as a productive member of society, you're going to be expected to be facile with these tools, you know, let's not weaponize them. And I really feel like every degree, every technical degree, every liberal arts degree should include some component

of ethics. What's your perspective on that? I don't see that happening fast enough, but I feel like it's our obligation as the community training these students to introduce them to the principles of ethics. Absolutely. I think ethics is

probably should have always been and should always be a central enduring skill and mindset that is taught explicitly to learners. And that is especially true today with the moral and ethical dilemmas that AI presents. So I absolutely believe that's true and that's part of the answer. And then the other part of the answer is,

as I was sort of hinting at is the very nature of assessment can and will change with AI powered tools.

It's not going to be, we don't need to have high stakes assessments based on steel trap memorization of facts into the future to understand how a person is learning, what they've learned, their level of mastery, right? We're going to have new tools that not only allow us to do that, but allow us to do that all the time.

And they provide those insights back to learners in ways that are motivating for them so that they don't need to cheat because it's not about like jumping over some hurdle. It's about embracing learning

every day, all the time as the central, central part of why you're in school or why you're getting a credential. I think AI tools can actually help us to do that better. And we've certainly seen that at Coursera as we've put some of these tools on the market and have been testing them out with learners who are responding very positively to them. As prep for this conversation, I listened to a great conversation you had recently where you cited the impact that using AI to translate

courses into multiple languages at probably higher level of quality is having on the learning experience. I think related to the point you just made, can you maybe

Share that example. Yeah, that's a really great example. So, you know, three years ago, two years ago even, it was very expensive to get to sort of authentic, you know, trustworthy people.

translation of content. And for us, having now over 170 million learners worldwide, that was a problem, right? Because really we were only providing access to learners who could

access our content in English and access our highly academic and professional content in English, right? And so the AI-powered, machine-learning-powered translation capabilities that have emerged over the last couple of years have totally turned that reality on its head.

So that not only are we able to provide translation of all of our didactic content, like readings and videos, but we are also able to provide translations of all of our activities, all of our coach interactions, dialogues going on between learners and our coach tutors. And

And they can not only see it and hear it in their native language, but they can toggle back and forth, which is tremendously powerful. We're hearing from learners who are in our certificate programs and our degree programs that that's exactly what they're doing. Like they want to learn it in Spanish or they want to learn it in Portuguese, but they want to be toggling back and forth between Portuguese and English so that they're developing sort of that connection.

proficiency in both languages. And that is like, that's a dream come true. Yeah.

Next phase of that, which we have been piloting in a subset of our courses, is the AI dubbing. Because a lot of learners, it gives them access, but they're not too keen on reading subtitles, for example, for the videos. But we're really seeing that with videos where we're doing this AI dubbing, which is not only more engaging for learners, but it's getting better and better all the time because it's not just the dubbing like you could imagine the old like.

I don't know. I remember when I was a kid, the old Godzilla movies where they would dub things and the mouth was all out of sync with them. These are like super synced, but they're not only super synced, that voice quality is there of that initial instructor or that initial professor. And those mannerisms are there as well. So that just makes it so much more engaging. And as you say, it's having a big impact on engagement and persistence.

both in just our regular courses as well as our certificate programs. And that's just, that's powerful. And it's a great example of something that was not even in our reality two years ago and now is like just a common tool. And it says your language and your culture are first-class citizens. Yes, yes. Everyone should have equal access to learning. Equal access. And I do have to say this, like,

Right now those tools work really well when the base language is English and you're translating into other languages But I do look forward to the day when it works just as well when your base base language is Hindi or your base languages is Portuguese or Japanese because like We in the US and in the English speaking world need access to those perspectives as well in the same ways and

As we talk about all the benefits of AI as part of learning experience, it's also our responsibility to talk about potential downsides. We know that AI is, at least when we're talking about language-based AI, it's based on statistical modeling, sometimes gets things wrong. It certainly can replicate human bias. What do you think is our obligation as we're

Teaching learners and introducing more AI into the curricula to also educate about what could go wrong. Yeah, I mean, for sure, AI can widen rather than close the equity gap, you know, if we're not super intentional about this work.

For all the bias reasons you mentioned, as well as if the best tools are only accessible to the wealthiest institutions or learners, or the best integrated curricula is only available to the wealthiest institutions and learners, will reinforce the very barriers that it's the promise of this technology to really break through and

So, you know, from our perspective, that's why I'm at Coursera, because we're scaling affordable housing.

AI-first, increasingly learning experiences and platform-wide capabilities across 175 million learners in very affordable, accessible ways. And I hope that we're creating models with these over 350 content partners we're working with, many of them universities,

that they can try out and experiment with on Coursera and then take back to their institutions and use there more broadly to support lots of different educational models. Well said. Marnie, I've got to get you off the hot seat, but you're not going anywhere without answering one last important question for me.

So roll back the clock to maybe Marnie the kid who was a gaucho at UCSB or Marnie the grad student at UPenn. You studied educational systems. It's your PhD. What has changed? If you were to get a PhD today in that same topic area, how would that curriculum be different? And what would it mean to have the same degree but today? - That is a really good question.

And I have to say, you know, when I got my Ph.D. at Penn and I love Penn and I am ever thankful for the education that I got there. At that time, it was pretty difficult to to really focus in on the integration of technology and technology.

curriculum and learning and learner progression. It was really early days in all of that work, and it wasn't a huge interest to education schools across the United States at that time. And I think that's changed a little bit.

but frankly, not enough. We're at the precipice in education of being able to, this is going to sound nerdy, but being able to understand the relationship between a learner profile, intention, motivation, sense of belonging, their engagement with different learning interventions, their persistence as a result of that engagement and the outcomes that they have.

We're at the point where we can understand that. We will be able to understand that better than we have ever understood it before. And with that understanding and that insight, take action in the development of new models. And I wish there was a...

a PhD in education and undergraduate program in education at every university that was just focused on that. And the learner analytics that really drive our understanding of that, it's important and it's a real opportunity, I think, for ed schools, engineering schools, social science schools into the future. And I hope they will move in that direction. And I would love to go back and take one of those programs again.

Well, this discussion, this space is evolving so quickly. We may have to have you come back and talk about the next version of this conversation because a year from now, I don't even know what the same conversation would be like. We can't know. That's the world we live in. Like I said at the beginning, I think what you're doing is going to indirectly or directly impact every human. So I'm glad someone like you is thoughtful about how we approach these things. So really good hanging out, Marnie. Thank you. Great hanging out with you. Thanks for having me.

Where can the audience learn more about you and the work you're doing? Well, I would encourage the audience to come and try this stuff directly by engaging with some courses on Coursera where they can experience the translations. They can experience Coach. They can experience some of these cool activities and assessments that we're developing that are powered by AI. I would love if they would do that. And we always welcome feedback through those channels. Brilliant.

All right. Well, gosh, this one sped by, but that's all the time we have for this week on AI and the future of work. As always, I'm your host, Dan Turchin. And of course, we're back next week with another fascinating guest.