Loading summary
A
This session was recorded live at the 2026 ASU GSV summit in San Diego.
B
I'm excited that we get to do this conversation because I think it's been coming up but it's in some ways addressing a missing link. Workforce and education have traditionally I think been thought of as yin and yang of supply and demand where skills in a skills marketplace and education and I'll count myself in that category by the way, I should say, I guess Ray Sass from Lightcast Lightcast is a labor market data company. We work with private sector, public sector and education entities to drive labor market intelligence into business decision making and a former educator myself, prior role was at University of Florida where I led strategy and innovation and former business school dean. And so I've been in that space a long time and public federal policy roles as well. But we think about supply and demand in skills context and education is producing skills and the private sector and public sector are consuming those skills, are using those to drive the marketplace forward and drive the economy forward. But we've seen a tectonic shift where those two realms seem to be drifting further and further apart. I want to introduce our very accomplished, impressive panel of folks here who are working on what I think of as potentially the bridge that could at least relink, if not bring those two back together. That is using work itself as a medium to connect work and education. Oh good, you have titles there. So then we won't do that. If you can, you can see those. So we'll make the most use of our time. But a couple of data points from the lightcast perspective that I'll offer as stage setting one in 2021-2024. Over those three years we saw the fastest yet pace of skills change inside of jobs. We saw the median job experience about a third of skills change. But one in four jobs saw 75% of the skills required to do those jobs completely change. That's an unbelievable overhaul in a skills landscape. And that three year change was roughly equivalent to the prior five years of change which was roughly equivalent to the prior seven to eight years of change. So a massive acceleration of that change at the Same time in 2026 is currently projected to be the worst entry level job market since peak Covid. In 2021 we are seeing roughly 26% increase in applications and about a 16% decline in job postings for entry level positions. On the backdrop of that speed of skills change we know it takes about an average of 30 months to construct a new degree from conception to releasing it to the marketplace. And when you think about what's changed over the last 30 months, you could say quite a lot. When there was a time about 10 years ago when I'm embarrassed to say that I was part of strategy sessions thinking about how we could build and launch coding boot camps. Who remembers those, by the way? How many educators in the room? Okay, so no offense if you did that. I hope it worked out. But we've seen a lot of change. If you're thinking about designing a degree program today for the demands of today, you may end up disappointed by the time that's completed. That's a trend. Enter apprenticeships I want to talk a little bit about the data there. Over the past 10 years we've seen apprenticeships double the number of them available. We got about 680,000 registered apprenticeships. It's really hard to track everything that would qualify as an apprenticeship, but we've seen about a doubling and there's been a little bit of denominator change on what that means. But it's about one apprenticeship for every 30 enrolled student in a degree program. That's a directional improvement. It's a market signal that people are looking at work as a way to link to education and improve the relevance there. But. But it's still a far cry from being ubiquitous. So I think I would say that the other big shift we saw in 2025 maybe final point I'll make on data before turning to our panelists to hear about the data that they use to guide their work is we saw higher ed in a rock and a hard place situation, as I would call it, with a bit of a rejection of the degree premise. And we saw 70% of open job postings no longer reference a degree at all as a require. And that's not just about blue collar growth. It's also an erosion of jobs that used to reference degree requirements in white collar spaces and a massive decline in those. And we saw overhaul coming for accreditation. We saw something to the tune. Hard to measure all the indirect costs, but something to the tune of $300 billion of investment in the higher ed sector from federal sources begin to leave the space over a 10 year projection. Some others might have better or additional insight on what that magnitude of that shift has been. But all of that to say education is looking at a world that's saying hey, we're not so sure that this is valuable. It was kind of wild to me to think about that last year I was a part here at ASU GSV of a lot of conversations about skills as an idea that we need to be focused on skills. I haven't heard anybody talking about that this year. I think we're beyond that piece. The question isn't do we need skills? Do we need to be focused on skills? The question is how do we build them and how do we communicate them? And so with that in mind, I'd like to turn to these four folks who are working on that problem. And I guess since we please do say your name and your organization, but let's go quickly through what is data that you focus on in your work and then we'll come back with a couple of other questions. Thank you. Me to start, please go right down.
C
Hello everyone, I am Stephanie Veeck. I'm with the American Institutes for Research or air, and I'm specifically in our economic opportunity area, including employment. But we cover the whole realm from early childhood all the way through lifelong learning, including supporting people with disabilities and communities. I will be talking from the perspective of as a national intermediary for registered apprenticeships as well as the work that we do to help serve as an intermediary of intermediaries, cross sector partnerships and a study that we are currently conducting. My colleague Samia Amin and I are leading in understanding the critical role that intermediaries play in scaling economic mobility through intermediaries. And what is it going to take? We know that there is a critical need for many to many transactions. There also has to be a lot of many to one. But at the end of the day, a placement in a work based learning opportunity is a one to one transaction. And if we can't figure out how to identify the critical roles of intermediaries and how they get from big pipeline of employers and pipeline partners, whether that's schools, colleges, adult training providers, into the relationships with one on one employers and that specific placement and working with those pipeline partners and employers for that placement and funding those and thinking about how that works, we're just going to keep having the same problem. I'm also coming to this from the from a perspective as an entrepreneur, an employer, a leader in workforce development, a leader in economic development. And these problems that we're talking about aren't new. We've been talking about these same dang problems for 30 years or more. And we know the answers. We just choose not to do them because they're hard. And it's time to stop avoiding what's hard and focusing on what works because we know it works. And realizing that business and learning and skills are changing every single day now. So we don't have time to talk about it. We have to figure out how to do it. And so that's the perspective I'll be bringing.
A
So the math on tuition is broken and enrollments are under pressure. Why is this? Because the perceived ROI isn't there for higher ed. What's the number one thing that institutions have to figure out? It's actually how you graduate people with work experience. Why is that? That's what companies care about. So we talk about like what? What is it? What, what we know. We need skills. So what does that mean? It's two layers. The first is can I market myself? The best way you can market yourself is have real experience. Because a company is looking at what's on your LinkedIn, what's on your resume, not what's another badge, another credential. They don't care, right? I mean, maybe it's a check the box of your degree, but they care about, okay, what did you do? And then it's proving the work. Okay, what did I create, what did I build? Did I make an agent? Did I do an investment memo? What can I tangibly show that I know that you didn't just have AI write that line on your resume? And my AI screener tagged it, I don't have time. Show me the proof. So that's where we're heading. And Northeastern, I think has a 5% acceptance rate now because of the co op program, they can guarantee work experience, but you can't build a co op program in a year or two. That takes. You gotta really build an institution from the ground up. So the question we look at, I'm founder of Extern and what we've been studying with this is how do you make a work experience that is taking the workload off the company? Because we're expecting companies to take on so much. You should be taking more internships. Why aren't you taking more internships? Well, they're not going to take, they're going to do what companies do, which is, you know, capitalism.
D
Right?
A
We can't change that. So we got to incentivize people at companies to take on more people. So our model is one to many, cohort based, where we take all that workload off and then we've placed now about 70,000 students into these externships as we call them, that we run for companies like Amazon and TikTok and startups and then make sure you have the training and the outcome in the portfolio that you can go and recruit with. And our data has shown that 74% of the externs who complete will go on and land a job or an internship in the market within a year. And it's also the confidence, the boost that they get when they go into an interview. So anyways, those are some of the data points we look at.
D
Hi, everybody. Heather Stefanski. So I lead learning and development for McKinsey. And the lens that I'm bringing to this conversation is about how do we actually design the work to be developmental, right. And how do we think as an organization, all of our organizations, to think about, how do we do this? And we talk a lot about formal apprenticeship programs. Well, at McKinsey, we have a culture of apprenticeship period. And so to me, as organizations, we need to be able to do a better job of teaching our folks to be more intentional learners and teachers in the context of the work that they're doing. Because I think a lot of the situations you talk about these internships, that two things, One is it maybe it's the companies don't want to invest in it because it costs them money. But the other piece is they may not have people in their organization who are actually qualified to help to teach these young people to grow and develop. Right. And how as we as organizations can actually think about that as a valuable skill. And by the way, the thing that's most exciting about there's actually benefit for the company because if you can create an organization that is an intentional teaching and learning organization, talk about the ability to upskill at scale fast. I no longer have to pull people out of the work to send them into learning programs to teach them. They can learn as they work. So it's a really powerful thing. So I've been doing a lot of research at McKinsey about what does it take to create this apprenticeship culture? What does that look like in an organization? You know, we all know that if you have a better teacher in the classroom, guess what, you have better performance of a students. But how much time do we actually spend in corporations teaching people to be effective teachers? That's one lens. And I would say the other lens is then you layer on AI and AI's role and their ability to actually support that teaching in a company. We think a lot often about AI for productivity's sake, but AI can be an incredible teacher as well. And again, how do we support that and build that into the apprenticeship model?
E
I love that. Good to be with you all. I'm John Collicker, CEO and co founder of Leland. We work with a lot of employers and schools and students directly to help them land jobs or level up in, or you know, train themselves in AI or other, other skills. I think the, you know, there's the data around a wage premium for people that, that know how to use AI effectively and that, you know, the data is anywhere from 40 to 60, 60%. Right. So it's a pretty big wage premium. I think the, the perspective I'll offer is sort of seeing what that actually translates to in sort of like a real tangible way. And I think on one hand of the spectrum you have entry level talent that is AI native and is using these tools in a way that makes them five, ten times more productive than a typical entry level graduate. They are bringing such a unique perspective to their employers, often training the employers themselves on how to use tools that they are being fought over by companies. And then on the other hand you have entry level talent that is not AI native. And so they not only need to be trained in the industry and in the job and sort of like build career judgment, but they also need to be trained on how to use AI. And so they've never been more worthless to an employer, which is unfortunate. Right. And so you end up with this polarizing spectrum. It's actually a massive opportunity for the class of 2026 because if they can learn these tools, you know, companies are literally fighting over them because I mean we, we, we had an intern on our team that we gave a full time offer to and he got like six full time offers from all these different places because he had built so much, so much like in the time he was interning with us, he had built all of these different systems and agents that were all talking to each other and it was, it was so clear that he was going to be teaching, you know, 15 year engineering veterans on our team, which is such a unique phenomenon that doesn't happen very. Like it's very rare that you would have an entry level talent bring a skill set that would be valuable to industry veterans. So it's actually a massive opportunity. And so I think there is a lot of hope in this. There's definitely a lot of change, a lot of nervousness, a lot of challenge. But I think if we can figure it out and you can get your institution and your students to adopt some of these tools and then the organizations are set up well to continue to build on that, it can be a huge advantage.
A
I'll just add to that another stat handshake shows. Okay, internships are declining, entry level roles are declining, but entry level roles asking for AI skills have 5x'd in the last two years. So this is the hope for anybody. But there's a lot of talk about young people coming out of college not being able to get jobs. If you find that river that you're talking about, you will leapfrog the job market.
B
Maybe we stay there for just a second to say one of the challenges that we see now is there's an entry level job challenge. There's really a middle 5 to 7 year of experience problem that we have. And one of the things that I've heard in many places is if you aren't doing the entry level work that you struggle with, what you develop by struggling through some of those tasks that are amenable to automation and outsource to technology, is you don't develop the judgment that gets you to become someone who has a different type of skill that's harder to train without having done the reps in the first place. So maybe I'll throw that to the panel of do you see that as a problem that will solve itself? Because things are changing so fast that what is required in a 5 to 7 year experience role 3, 4, 5 years from now will have changed as a result of entry level work being largely leveraging AI to get to there? Or is that a problem in the pipeline where we've removed a rung of the ladder that's necessary to get to more senior level roles later?
D
I mean, at least for us, I would say it depends on the skills, right? So for example, everybody, we really excited if AI created PowerPoint charts, right? But I know that I need to have distinctive problem solvers and I know that McKinsey is only going to be able to be distinctive with our clients if we can actually solve the problem that AI can't. So those skills I care a lot about and I'm being very thoughtful. And what we're really trying to do as an organization is think about how you design again, how you design the technology to teach, right? So it's not replacing the problem solving process, it is enhancing the problem solving process. And as a result you can actually get more reps on the thing, right? Because you can actually do it more times more quickly and you can get more reps if it's designed, correct. So an example is we have, we've just rolled out this new problem solving support tool. People may have heard of McKinsey's seven steps of problem Solving, right? You start with a problem definition, then you build an issue tree, you work through these seven steps and that tool helps you do that. Right. That tool also forces you to make sure you take each of the steps because oftentimes our client servants teams want to cut corners and they want to jump through something. But actually I'm going to force them, I'm going to force them to really think hard about a problem statement. And we actually know for a fact that teams who define a better problem statement actually are happier teams because they're not spinning their wheels. And we get to better client outcomes from the taking that very first step and doing it very well. And so if I can build a tool that's going to help people actually create a better problem statement and sort of teach them to do that in the process, I'm excited about that. Right. But again, I think we have to really think about what are those skills that are going to make our organizations to truly distinctive in the future. And then those are the ones you have to make sure you're not losing that sort of first or second rung on. But there's a whole bunch of other things that like
C
I want to add to that from air's perspective as an employer. So I know I said I was here as the intermediary voice, but AIR had no choice but to do some major layoffs last year. As we are lucky enough to get to rebuild back up over this last year, we've had to retrain ourselves the same. We've all had to learn to use AI. We've all had to learn to use AI smarter, to get better at figuring out what the right prompts are, to figure out how to have that critical eye to using AI effectively and responsibly and not substituting our expertise with AI. And as we are getting ready to put out job recs, like up to 100 job recs, they aren't for the high level positions they, they are for our first and second level. And the biggest challenge is how do we take everything we've spent a year learning that we're taking to the market to help our clients understand how to use. And how do we make sure that we are hiring the right people with the right attitude and aptitude to learn AI alongside us instead of being just another one of those employers out there in the marketplace trying to hire the few people that already have it all? And I think that's a challenge that we all have to think about as employers, is that not everybody is going to have that, but you can figure out how to take what we've learned rapidly, what we've all learned rapidly, and build it into the way we train and so I love what you're doing at McKinsey. That's exactly what employers need to be doing. And as we're thinking about the role of intermediaries in that space, it is realizing that the first job of business, as you said, Matt, is to run their business. It's to make a profit, it's to exist and to deal with today's critical challenges. Whether you're a for profit, a nonprofit, you have to do today's business first. But if we're thinking about how are we helping people get that real work learning experience at work, in the classroom, about work all the way through the work based learning continuum, then I think we can get smarter about building more of the amazing people like John mentioned have the ability to go in and lead up and to manage up and to teach all of the rest of us that have been around too long how to use AI better. And I want to just really quickly acknowledge something that John said. We got to work through our intermediary project with an amazing man who was on the spectrum, had gone through two boot camps, had a college degree in computer science, was. Had been working in video editing because it was the only place he could get a job. Finally quit because he just couldn't stand it. Went to an apprenticeship job fair and got connected with this incredible guy at Miami, Ed Tech, who offered him a job as an AI generalist, not even knowing what it was, but knowing that two years ago they needed to figure out what this role's going to be. The very cool thing about Norbert is he took the job on, he said, I can do this. They started out with apprenticeship standards that they thought were the right things that an AI apprentice would need. And through the process of doing the work, they completely flipped what the reality of an AI generalist, an AI specialist, was. And over that two year period of time, they rewrote their apprenticeship standards twice. And Norbrecht the apprentice, was the one rewriting the apprenticeship standards because he knew what it took to do the work. And that's where I feel like we need to build more of that into the working, working and learning together. Whether it's an experience or it's part
B
of the job, that's great. I think it's a really good segue to two other related topics. One is, if we're talking about what employers can do to build the skills that they need internally, let's flip it for a second to say, how should educators be thinking about bringing work into this space? Or we did vice versa. How do you think about bringing work into an education space. And then I will move quickly after that to how do you prove that that actually happened? When you're not the employer that's building the skill? When you're building the skill internally on the projects, you have the project as evidence that it happened. When you're building the work in an education environment, what is a meaningful signal that skill attainment has happened? So maybe I'll go to John and then Matt.
E
Yeah, I think it's so we've all known for a long time that movement doesn't mean progress. And I think we've all maybe worked with people that are always busy, but it doesn't feel like it's connected to real impact. And I think that the interesting thing about AI is that it accelerates movement dramatically. And so regard like the direction, you know, we've talked. There's, there's kind of the, you know, you want to hire barrels, not, you know, ammunition or, you know, there's like a concept of like bricklayers versus the architect. Historically, you actually did just need people with the skills to go and execute. Well, turns out AI is really good at doing the execution. And so the architect or the barrel or the judgment behind it all matters so much. And so there's the initial skills on how do you wield AI and how do you train that. And I think as educators, we all need to figure out how do we, you know, build our own curriculum or we obviously offer curriculum as part of our programs, but there's lots of different options. But then you also can't forget that the judgment behind it all has become more important than it's ever been. And there's this term slop cannon that is, that is thrown around because with AI producing, you know, so much content or so, you know, so much code or so much, you really have this problem. Like, my biggest problem with my team right now is not production, it's prioritization, it's direction. It's making the hard calls. And so I think, you know, there's. We can get into maybe some of the tactics for how you actually train these skills. But I would just offer that the actual skills you are building, I think, you know, are role dependent. But there's this element of wielding AI and then there's this element of high quality judgment. And students are going to need that much, much faster than they ever have. Because there used to be jobs for them to come in and write code and be told what to do. Well, now the code's being written for them. So if they can't write the code. Well, you know, if they can't write the code, which is the AI, you know, AI doing it and they don't know, you know, what to do or how to fit in and the judgment behind it, it's really hard for them. And so I think I'm just going to offer that,
A
I guess in terms of how schools should be operating. Going back to earlier point of I think any institution next three years as the economy I think is going to be pretty weak. You're going to have to figure out work, integrated learning. It's going to have to be a foundation of what you do. And there are some that are moving very quickly and they're like some of the smaller schools or departments that we're bringing AI and we're just diving right in. We're not waiting for the academic senate to go off for a year and figure out what we should do. And those are the right institutions. Now what are they doing? Doing as much as possible. It's bringing the companies in and having the work be as real as possible and as less of a simulation and case study as possible. And there's like a gradient of that. The problem is that's really hard to do and it's too much time. And so this is where yes like AI comes in and like. And what we've done is we've, we ran these externship programs with a lot of human support first. So we would create the curriculum for the company. They would tell us here are the goals, but we don't have time to create training. And then we would help the company would come up and do a webinar with the students and answer questions and share what they're doing. But then we would have support staff answering questions. And what we've now done over time is we've started building AI agents to sort of at this moment it's like humans are still in the loop. They're still qaing, but they're now the subject matter experts for the students saying hey, I'm stuck at midnight. Am I working on this intrusion detection for cybersecurity externship? Am I going about this the right way? What's some feedback? Can you help give me an assessment? How would my manager view this? And that way we can help give them more reps faster, help them practice over time. Judgment comes from reps and mistakes and help them make those mistakes. But in a low risk environment, not where a manager is like the student doesn't get it. I don't have time for this anymore, which is kind of what the default is.
D
I mean, that's the point. The managers don't have necessarily the skills, and the AI can help provide the internship. But I think this question of what markers to look for that you kind of raised.
B
I was just going to ask you.
D
Yeah. Is a really great question. Right. So if you think about more traditional. McKinsey. We cared a lot, let's say, about your problem solving skills. So there's markers. You looked at grades in certain courses. You could look at SAT scores, you could look at GMAT scores. There's some things that we could look at. Right. But as we're starting to care a lot more about the human skills, because again, to your point, AI can do a lot of the computation. So maybe I don't really care so much about your quantitative skills. I mean, I still care, but, you know, and I'm starting to look for sort of more of these, like how. How good are you at building relationships with others? Right. And so what you're describing is an environment where those students are actually getting the practice to sort of build the relationships. But then how do I know if somebody's really great at building relationships? Right. And in the past, you know, we. We might have a set of academic institutions that we knew that if the person was the president of the student council at such and such or had this particular role, they were going to have to have build relationships. So that was kind of a signal to us that they build those relationships. But that's really narrow because we can only know about so many schools that can know so many things and sort of build some relationships. The other thing that's a really great marker is we do behavioral interviews. And in the course of a behavioral interview, we can really understand that also doesn't really scale. Right. So we have had in the past this game called Solve, and it's actually a video game that students play so we can understand their problem solving skills. And we actually used to have a problem solving test, and we actually moved away from that because we wanted to make something much more accessible to many more people. And in this problem solving game, you don't actually have to know any business concepts. It's all about, like, creatures and building an ecosystem and all these things that students actually say it's kind of fun to do. And we're assessing their problem solving skills. So we are actually about to launch another video game that's much more focused on behavioral skills. Because again, our goal is to try to be able to assess behavioral skills at scale to Give us some markers or indication. So I've forgotten what it's called. It's got another cute name. It's not Solve, and it's going to be launched in the next six months. Because again, we care a lot about those skills, particularly in the world of AI, when the human skills are going to matter most and they're also much harder to identify. Maybe one other thing that we also know, but again, it's a little uncomfortable for some, as you guys are doing those AI support and coaching, et cetera, the AI is actually also capturing data on how well you're doing those things. In our organizational context, for the people that we actually have employed, we are implementing much more deliberate practice into sort of all of our learning experiences. And as a part of that deliberate practice, I can then kind of AI can help me grade people on how they're doing on relationship building because they're doing a role play. And that role play, by the way, can give the person feedback on how well they're doing it. But then at an aggregate level, I can capture all that data and I can say my people are getting 20% better at building relationships over the course of onboarding because they've done these deliberate practice exercises over time. So I don't know the answer to your question exactly, but I do believe that we have the ability now with AI and with technology to potentially do this at scale, in a way, on some of these other human skills that we've never been able to do before. Because we have the ability for kind of AI to support kind of assessments in a way that aren't tests.
A
We're doing something similar just very quickly. So taking the portfolio of what you've built, being able to assess that, but also all the interactions during the externship, like, did you ask questions in the live session that we recorded? Did you meet deadlines on time? Because we can now monitor all this on our platform. So we want to create this profile that you own the data, but you can go to an employer and it's like, you know what, someone who is engaging and struggles, but comes back and asks good questions, like, I actually want that on my team.
B
Well, one challenge there is. Stephanie, I've heard you. Excuse me, Heather. I've heard you say that in the expansion of McKinsey hiring, you can't get enough talent from just the traditional elites where you happen to know the curriculum inside of Harvard. And so you know that this course and that grade means a particular thing. So if you're hiring from 1,000 institutions instead of 20. What are the signals that those institutions, the markers that you mentioned, how can they signal that something's happened where you don't have to learn to your point, you have to learn what the internal process is of how I'm building something like the credential. What is the marker that maybe the way. I'll say this is advice for educators on what they can do to signal that real stuff is happening. Because I think that we more or less could all agree that grades are not a very good indicator of what's happening inside of a classroom, nor is the degree when you say, well, an MBA. MBAs were very fashionable 20 years ago, and now they're saying, well, I don't really know what that means, so I'm going straight to a resume of what skills do you likely have behind that? So, interested in your thoughts on that?
D
Yeah, I mean, there are, like, some things that we're, like, hiring more athletes these days, right? Because, like, you know, athletes, we have team experience, especially successful athletes, right? They've had team experience. They've been coached along the way. They've had to work really hard, like in school and with academics to sort of get their way through. We know they've had struggles, right? So, like, that's like an interesting marker for us. So, I mean, you know, something, you know, something along those lines. You know, folks in the military is another, like, source that we're kind of leaning in more to, because we're trying to find places where we can find folks with, like, grit, teamwork, et cetera. But this is again, part of why we've launched these problem solving games, this problem solving game, and now this behavioral assessment, because it allows us to actually get a whole bunch. I mean, like, I can send millions and millions of people, I can have go through this because there aren't great markers. And I would say the other thing. A lot of small universities and institutions, we actually have to rely on the professors and the leaders of those schools where we'll call them up and say, who are your best? Right. And then they, like, handpick their best and send them to us because otherwise we don't have enough understanding. And they're really, again, other than a few things that are common across universities, it's tough.
C
I would add that I think it's really important for educators to recognize that you are part of an ecosystem and you don't have to do it alone. So your intermediaries, your workforce boards, your chambers of commerce, your trade associations in your local area are, are all trying to solve this too. And by working together to do that, you can do that better. In our interviews that we were doing in Indiana, we interviewed superintendents, CEOs, trainers, all sorts of folks across the state of Indiana. And one of my favorite examples is a small county in Indiana where the four high school schools and nine employers, they don't have any big employers, they have medium sized manufacturing tech employers. Nine key employers and four high school superintendents came together and said none of us have this capacity to do it alone. So they identified the trusted partner between all of them who could they all agree that they trusted and they used that trusted partner. In this case, it happened to be an education alliance that the schools trusted, the employers trusted, they came together and they created work based learning experiences that were modeled off of project based work, were similar to co op or internship models and could evolve into a full blown apprenticeship. A lot of those employers have apprenticeships. Some of them are Swiss and German companies that want people to grow into apprenticeships. But instead of saying, saying we're going to have one solution for every employer in our area or we're going to work with one high school, they use that intermediary model to come together and say what is the project based work that this group? And they have an advisory board of nine employers across four different industries that are saying what are the critical skills we need and how are we going to get involved in your classroom so that we can help you figure out how to assess whether the students are getting those. They are going into the classroom and doing the interviews and having the students ask them the questions and they're dedicating staff to do that. It's way easier than trying to bring 20 cohorts of 20 or 25 students into their business to do it. They're sending leaders into those classrooms to do it and they're identifying this next steps for those internships and those apprenticeships while they're doing that. And that's just that role of that intermediary. Again, it doesn't matter who the intermediary is, as long as they're trusted by both the employers and the pipeline partner, whether that's K12 or colleges or whatever. But we can't just keep having everyone try to do it themselves and think that the employers are going to jump at that chance.
B
It's really good. You actually I'm going to do lightning round. We have three minutes left. So one for each of you. I was going to do a summation question that was sort of a how to start. So advice for folks on how to start. And I think if. Tell me if I'm distilling yours wrong, but it would be, don't do it alone and get to some partners to help you do this. Maybe that would be. Maybe I'll just go down the line advice for either educators or employers on how to start making a difference. If you bought into the concept that we should be doing this, but you're resource constrained, maybe you don't have the scale of a McKinsey to do in house. How do you think about getting started?
A
I mean, obviously this is self serving, but we built a platform to do this and be the intermediary. But you know, work with a company like Extern to handle all of that for you to tiptoe, learn from it and then start to create some of your own programs that, that you meet your requirements.
D
Maybe just first say, I'm super interested to see what Matt's profile will come come about. So Matt, we're gonna have to talk when you guys have that for your candidates, because I think that would be really interesting. And I would just say, you know, from an organization perspective, to start is just like really think about like, do I have teachers and learners in my organization and can I create sort of more of a constant flywheel? So if I'm bringing in raw, smart talent, I'm convinced that I can actually develop them effectively. Right. So I would just say, how are you valuing the role of a teacher in your organization?
E
Yeah, I love that. Yeah. I think there's a lot of things that people can do. And I would say leaning into your strengths as an institution is so valuable. And I think trying to connect to the physical world wherever you can in a world that has a lot of noise digitally. And that could be bringing employers into your institution physically, getting your students to their companies physically. And it could be, I guess, digitally in terms of a connection, but there still is a human connection. We posted a job and had 5,000 applications in 24 hours because they were all custom and they were all custom written by an AI agent. Right. So where there's more noise. The only way that, the only way you're going to stand out is by connecting to the physical world and leaning into the strengths that you do have and the industries that your students are already naturally interested in. I think Heather has done an amazing job of identifying which skills matter to McKinsey. And I think there's an assessment that every institution can do for their own people. Which skills matter for our students to land jobs here and you know, show don't tell is is the signal here. And whether that's being an athlete or, or doing something else, I think you have to show in a, in a world where there's just so much noise, really great.
B
Well, thank you all. I think we landed here right on time. So thank you very much for the conversation. Encourage everyone to talk with them after.
Episode: From Crisis to Capability: Reinventing Entry-Level Work
Recorded: May 5, 2026, San Diego
This live panel at the 2026 ASU+GSV Summit explores the crisis and transformation facing entry-level work, with a particular focus on bridging the widening gap between education and employers. The five panelists – representing labor market analytics, intermediary organizations, education technology, management consulting, and employer networks – discuss how accelerating skills change, the rise of AI, and waning faith in traditional degrees demand new models for both preparing and hiring entry-level talent. Together, they surface innovative approaches, critical data, and actionable advice for educators and employers striving to create relevant, scalable work-based learning in an era of disruption.
Speaker: Ray Sass (Lightcast)
Timestamps: 00:13 – 06:28
Speaker: Stephanie Veeck (American Institutes for Research)
Timestamps: 06:28 – 08:52, 19:05 – 22:50, 34:26 – 37:21
Speaker: Matt Wilkerson (Extern)
Timestamps: 08:52 – 10:31, 25:47 – 27:55, 31:30 – 31:59, 37:57 – 38:11
Speaker: Heather Stefanski (McKinsey)
Timestamps: 11:15 – 13:14, 17:08 – 19:05, 27:55 – 28:47, 33:10 – 34:26, 38:11 – 38:47
Speaker: John Collicker (Leland)
Timestamps: 13:14 – 15:41, 23:35 – 25:47, 38:47 – 39:58
"We’ve been talking about these same dang problems for 30 years ... We know the answers. We just choose not to do them because they’re hard."
— Stephanie Veeck, 08:00
"The question isn’t do we need skills? ... The question is how do we build them and how do we communicate them?"
— Ray Sass, 05:32
"They care about ... Show me proof ... not what's another badge, another credential."
— Matt Wilkerson, 09:28
"AI can be an incredible teacher... If you can create an organization that's an intentional teaching and learning organization... you can upskill at scale fast."
— Heather Stefanski, 12:03
"Entry level roles asking for AI skills have 5x’d in the last two years."
— Matt Wilkerson, 15:41
"It’s a massive opportunity for the class of 2026 because if they can learn these [AI] tools, companies are literally fighting over them."
— John Collicker, 14:31
"Grades are not a very good indicator of what’s happening inside a classroom, nor is the degree... Now [employers are] going straight to a resume of what skills do you likely have behind that."
— Ray Sass, 32:08
For Educators:
For Employers:
For Both:
End of Summary