What is Holding Back AI Innovation in Higher Education?
What is Holding Back AI Innovation in Higher Education?
What is holding back AI innovation in higher education?
Outdated regulations, especially those tied to seat time and “regular and substantive interaction”, are limiting innovation.
These rules were designed to prevent low-quality correspondence programs, but today they:
- Regulate how education is delivered (inputs) instead of what students learn (outcomes)
- Make it harder to scale self-paced, AI-enabled learning
- Reinforce faculty-centric models that don’t reflect modern technology
Why does AI require a new model of learning?
AI changes how people learn in two key ways:
- Students will start at different skill levels
- They will take different paths to reach mastery
This makes fixed-time, one-size-fits-all education models obsolete.
How much will AI change jobs and skills?
A major takeaway from the Capitol Hill discussion:
- 70% of the skills in a typical job will change within five years
This means:
- Nearly every worker will need to reskill or upskill
- Learning will shift from “once and done” to continuous and lifelong
Can the current education system handle this level of reskilling?
No.
Today’s system is not built to:
- Retrain the majority of the workforce at scale
- Support continuous learning for people who already graduated
- Deliver education efficiently enough to match the pace of AI change
What role should AI play in solving the reskilling challenge?
AI must be part of the solution.
According to Rajen Sheth:
- We will need to train everyone on AI
- And use AI to train everyone
That means:
- Personalized learning tailored to specific jobs
- Scalable delivery across millions of learners
- Education embedded into real work contexts
What did policymakers on Capitol Hill understand about AI and education?
There are encouraging signs:
- Federal agencies are already experimenting with scalable models (e.g., AI literacy via text messaging)
- There is bipartisan awareness of the need to support innovation
- Stakeholders across government, industry, and education are engaging together
However, coordination and speed remain challenges.
What needs to happen next?
To meet the scale of AI-driven change:
Policy must:
- Shift from regulating inputs → measuring outcomes
- Enable flexible, technology-driven learning models
Institutions must:
- Build systems for lifelong learning at scale
- Focus on skills, not just degrees
Technology must:
- Deliver personalized, accessible learning for every worker
Bottom line: What is the biggest takeaway?
AI is forcing a shift from:
- Static education → continuous learning
- Standardized pathways → personalized mastery
- Inputs → outcomes
The systems that adapt fastest will define the future of both education and the workforce.
Transcript:
00;00;05;13 – 00;00;30;24 Wes Smith As our audience knows, the President’s Forum was created to drive accountable innovation in higher education. The promise of innovation to improve higher ed by lowering costs and improving outcomes has never been more obvious. Our guest today is Rajan Sheth. He’s the CEO and co-founder of Kiran Learning. And last week we had the chance to cross paths in Washington, D.C..
00;00;30;26 – 00;00;33;02 Wes Smith Rajan, great to see you again today.
00;00;33;04 – 00;00;35;09 Rajen Sheth Hey, Wes. Good to see you as well.
00;00;35;11 – 00;01;05;25 Wes Smith Hey, one of the conversations that we had last week when we were together in D.C. with the President’s Forum, it revolved around some of the outdated regulations that are hampering innovation. And I just want to start with your reaction to hearing the president’s thinking around, rules and laws slowing down the adoption of AI and other innovation that could really improve the outcomes for students.
00;01;05;28 – 00;01;33;02 Rajen Sheth Yeah. Well, it was interesting for me as a technologist. It was a really enlightening conversation. I didn’t fully appreciate the barriers that universities run up against when they’re trying to do innovative things. What was interesting about that room is that you had the most innovative universities in the nation that were there, all who are open to innovation and who have been for many, many years.
00;01;33;04 – 00;01;52;12 Rajen Sheth But they’re running into barriers, with this. So, you know, I think one of the things that that really was interesting is when they started to talk about how there’s, how there is a, regulation on seat type, like, you have to be in class for X amount of time in order for it to be accredited.
00;01;52;14 – 00;02;33;28 Rajen Sheth And that is something that is just going to change rapidly in the world of AI. It’s just, you know, everyone’s going to start from a different starting point and they’re going to take a different path to getting to mastery. And that’s something that we’re going to have to plan for. With, with students that are out there. The other interesting thing that was interesting about the discussion is I came to appreciate kind of the coming soon that is coming with how every job is going to change, with, with AI and we’re going to need to make sure the regulatory environment is such that our best universities are most innovative universities able to educate those
00;02;33;28 – 00;02;37;17 Rajen Sheth students. And the scale of this is going to be massive.
00;02;37;19 – 00;03;07;09 Wes Smith Right, right. The conversation that you’re referencing, with the, the seat time, issues that that need to be resolved. It’s something that like insiders on, well, at least at the president’s forum are thinking about all the time. And it’s the regular and substantive interaction, regulation. The Department of Education was essentially trying to prevent, you know, low quality programs, from accessing federal aid.
00;03;07;11 – 00;03;31;03 Wes Smith And those used to be, you know, low program, low quality programs that were essentially, hey, this is this is, a program that you can send in. They’ll send you some material, you send it back, they’ll, you know, decide if you did well or not. And and they’re pulling down federal financial aid for programs that just weren’t very good.
00;03;31;05 – 00;04;05;01 Wes Smith And so it had a great intent as it started. But but it started regulating, the the inputs rather than the outcomes. And when you start regulating inputs, well, when the inputs become, more effective, delivered through technology, well, then your regulations are outdated. It kind of hard codes, faculty centric models and and it really hampers, you know, self-paced learning.
00;04;05;03 – 00;04;06;15 Rajen Sheth Yeah. So yeah.
00;04;06;17 – 00;04;12;16 Wes Smith Yeah, that’s the issue that we have to deal with. It really discourages technology enabled scale is what it does. Yeah.
00;04;12;18 – 00;04;33;06 Rajen Sheth You know I think it’s a really good point that yeah, it’s almost like you need to separate the of the intent from the execution. And, you know, in technology, a lot of times for particularly product management and technology, we try to separate out the what versus the how. So like for example, when we specify a product, we try not to specify the how too much.
00;04;33;08 – 00;05;00;13 Rajen Sheth We try to specify the problem and then let the engineers figure out the how in the most creative way that that that meets that requirement. And I think we got to do the same thing with policy. We need to understand the intent and then the how it’s going to change rapidly over time. But then, you know, as long as we’re going towards that intent, which is a strong good intent, then we’re going to we’re going to be able to meet it in different ways.
00;05;00;16 – 00;05;08;17 Wes Smith Yeah. I mean, in this case, we’re we’re functioning under regulations that were designed for correspondence courses.
00;05;08;22 – 00;05;09;05 Rajen Sheth Right.
00;05;09;06 – 00;05;43;18 Wes Smith Exactly. And we’re so far past that, the idea that you that you’re regulating that the how it’s done, you’re going to be perpetually, you know, behind with technology that’s just the fact unless you can figure out, you know, the, the why, the why is going to be really important. Okay. Exactly. So after our meeting, last week, after the president’s four meetings, you joined an effort to provide, insight to Capitol Hill staffers about the evolving technology, specifically AI.
00;05;43;20 – 00;05;50;01 Wes Smith And, and that was up on the Hill. Can you tell us a little bit about the event and what you observed there?
00;05;50;04 – 00;06;13;17 Rajen Sheth Yeah. Yeah, it was a wonderful event. You know, one, it was just interesting for me to be in the capital. That was the first time I was I was over there. And you know, you are. You’re in a place where so much history has been made and you can see the, the, the kind of, the intent of that is there for so many people to kind of adapt where we are to the new environment.
00;06;13;19 – 00;06;34;20 Rajen Sheth What was interesting about this, though, the forum was basically a set of companies that had been thinking about AI skilling, you know, how do we scale the workforce about AI? There are a variety of NGOs and nonprofits that have analyzed different aspects of this. And then there was, there were government. So there’s the, congressional staffers who were there.
00;06;34;20 – 00;06;59;07 Rajen Sheth There was the Department of Labor, Department of Education, that there were all there. And it’s the right group to bring this together. What was really striking to me is to understand the magnitude of the tsunami we are about to encounter. And like Lincoln was talking about how 70% of the skills of any typical job will change over the next five years because of AI.
00;06;59;09 – 00;07;26;26 Rajen Sheth And the vast majority of workers out there are going to have to be reskill, for this. And we also talked about how the existing environment, whether it be, you know, institutions, you know, learning and development training within corporations, it’s just not scale to deal with that. We’ve never had a situation where that portion of, you know, that giant portion of employees need to be rescaled.
00;07;26;28 – 00;07;52;00 Rajen Sheth And, you know, what was really interesting for me and what I talked about there was, we’re going to need to train every single person about AI and how you use AI for what they do. And we’re going to need to use AI to to train every single person. Like we’re going to need to figure out how do we extend our institution so that it is personalized for that person.
00;07;52;00 – 00;08;06;24 Rajen Sheth That’s an accountant in this particular organization that’s now learning how to use AI or a, you know, a, machinist in a particular organization that’s trying to use AI. You know, those are the things that need to happen if we’re going to, if we’re going to meet that scale.
00;08;06;26 – 00;08;21;03 Wes Smith Well, this is when I hear that, from, presents for perspective, the, the amount of reskilling that we’ll need. Did you say 70% of jobs will need to be updated?
00;08;21;06 – 00;08;46;06 Rajen Sheth What’s the what he said was the 70% of the skills for the typical job will need to be reskill. And what that means is, actually, it may be even more striking than that. It could mean the vast majority of people are going to need to learn how to use AI as part of their job. In some cases, in small ways, in a lot of cases, in very big ways, in order to still do their same job five years from now.
00;08;46;08 – 00;09;20;28 Wes Smith So I mean, when I hear that, what I hear is the current system that we have will never accommodate that kind, that scale of reskilling. So we have to think through how we can reskill individuals in a much, much more efficient way. And we have to I mean, when we’re looking at this lifelong learnings, a tagline that a lot of, in higher ed of have been saying a lot, you know, you hear, oh, yeah, we’re moving towards lifelong learning.
00;09;21;01 – 00;09;43;02 Wes Smith This is truly one of those areas where we’re going to have to incorporate back into systems. People have graduated with what they thought were terminal degrees and. Yeah, yeah, we’re done. And we, and they have to come back and reskill. They have to they have to upskill. They have to learn more about. And we don’t we don’t have the system that can do that right now.
00;09;43;05 – 00;10;03;10 Rajen Sheth Absolutely. And I think what’s interesting is that over the past, you know, ten, 15 years, lifelong learning is, is has gotten more traction. But it isn’t nice to have as a person must have. Upskilling has gotten traction, but it’s a nice to have as the person must have. Now all of a sudden, it’s going to be a must have.
00;10;03;17 – 00;10;19;05 Rajen Sheth Like you cannot replace all those people with people that know it because nobody knows it, and everybody’s going to need to, is going to need to learn. We need to get our systems to the point where, yeah, where they can be able to train that volume of people.
00;10;19;08 – 00;10;44;20 Wes Smith Well, I’m assuming that this was kind of shocking, to, to Hill staffers and government employees to say this is the kind of, of massive change that we have to prepare for. Did you get any sense from them? About their, you know, their preparation for this or how they’re planning to, you know, facilitate innovation through policy?
00;10;44;23 – 00;11;05;14 Rajen Sheth Yeah. Well, I was actually very impressed with how much they are understanding what’s what’s hitting and the kinds of things that they’re thinking about. The Department of Labor, talked about some of the things that they’re doing. They demo, for example, a, an AI literacy, module that they, they put out via text messaging.
00;11;05;14 – 00;11;25;05 Rajen Sheth So you just sign up for a text message and then it takes you through a ten day course, which I’m taking right now. I’m actually going through it right now. And it is, it gives you kind of the basics of how you think about AI. In with my other. Had I told you this before? I teach a class on intro, the AI at Stanford.
00;11;25;08 – 00;11;46;12 Rajen Sheth And it was interesting because a lot of the same principles that that I focus on there things that they were teaching via text messaging. And so their point was you can reach many, many more people via text, and you can, you can get them the right information to make them not afraid of AI. And I think I’m impressed that the people are starting to think about that.
00;11;46;14 – 00;12;09;02 Rajen Sheth However, it’s going to need to be a coordinated, coordinated thing between the government, between institutions, between companies to really actually solve this problem. It’s it’s, you know, one of the biggest things we need to solve. The other interesting thing is that they talked about, particularly LinkedIn, talked about the economic positive impact that could happen because of AI.
00;12;09;05 – 00;12;23;05 Rajen Sheth Like it could be an additional $4 trillion, in terms of adding to our GDP. So it’s a huge amount, but we have to do it in the right way. To actually get there.
00;12;23;08 – 00;12;45;02 Wes Smith Well, I’m impressed just by the idea that the Department of Labor is ahead of the curve on this text messaging, campaign. That’s that’s impressive to me. I mean, if somebody is thinking ahead and saying, hey, this has to be a huge focus for us, we need to start educating and facilitating a transition to an AI world.
00;12;45;05 – 00;13;12;02 Rajen Sheth Yeah, absolutely. They have a guy named Taylor Stockton that, is their chief innovation officer and been thinking about things the right way. And, you know, he comes from the startup world. He was a Google before as well. And, I’m impressed that they’re they’re thinking in a very agile way. And, you know, obviously it’s hard to get things done in, in, in the political world, but I think we’re all going to need to work together to figure out how to how to move quickly here.
00;13;12;06 – 00;13;46;14 Wes Smith Yeah, absolutely. And the meetings that we had on the Hill with presidents last week, I thought it was pretty remarkable how consistent the responses were from, Democrat leadership and Republican leadership with regard to, facilitating innovation. I think both sides, they understand the issue and they want to solve the problem. The next step is actually, you know, putting some of the solutions into legislation.
00;13;46;14 – 00;14;01;12 Wes Smith That’s the hard part in DC, right, is getting something through and signed and, and, it’s just inherently political. But, on this particular issue, it seemed fairly bipartisan to me. I don’t know if that was your experience with Hill staffers.
00;14;01;14 – 00;14;02;10 Rajen Sheth Yeah.
00;14;02;12 – 00;14;04;06 Wes Smith We saw something different.
00;14;04;09 – 00;14;23;18 Rajen Sheth No, I saw the same thing. I think everyone is realizing that this is going to be an issue, and that is not polarizing. Everyone knows this is coming. And so, you know, we know we will have a big problem with the job market if we don’t do something about this. And so, there was a lot of unity out there.
00;14;23;18 – 00;14;39;25 Rajen Sheth And I think now we got to see what kind of policy can we put in place, what kind of innovation can be put in place, how that can be rolled out, and how you involve the institutions, and, and the other companies into this to, to make it a reality.
00;14;39;27 – 00;15;09;29 Wes Smith Right. Okay. Well, I want to wrap this with your takeaways. So based on your experience and what you saw, at the Capitol last week, what would your takeaways be for specifically for higher ed innovators who are looking to incorporate, AI and other solutions into their work so they can lower costs and so they can, you know, increase the outcomes, the quality of outcomes that students experience.
00;15;10;02 – 00;15;12;27 Wes Smith What are your takeaways based on what you learned?
00;15;12;29 – 00;15;37;09 Rajen Sheth Well, I think that a lot of what we talked about in the president’s form about bringing down the cost of education, reaching more people, it’s going to become vital over the course of the next few years. And I think that we’re going to have to move quickly. We’re going to have to adopt AI in our teaching practices and make it such that we can really, truly personalize for every situation, every learner that’s out there.
00;15;37;11 – 00;15;50;25 Rajen Sheth And we’re going to have to remove the barriers from a policy perspective such that we can all move faster. There’s a lot of intent to move faster, but it’s really hard to do that right now. And so if we can do that, we can actually meet this challenge. It’s about faces.
00;15;50;27 – 00;15;57;05 Wes Smith Yeah. Well Rajan, thanks for joining us today. Thanks for the debrief on the time in DC. It’s been a pleasure.
00;15;57;07 – 00;15;58;24 Rajen Sheth Yeah. Thank you. I really appreciate.
Recent Comments