‍

Moderator:

• ‍Michele Chang, Chief Strategy Officer, Strada Education Foundation

‍Presenter: 

• Andrew Hanson, Senior Director, Employer Alignment, Strada Education Foundation

‍Panelists:‍

• Alex Alonso, Chief Knowledge Officer, Society for Human Resource Management
• Ashley Finley, Vice President for Research and Senior Advisor to the President, American Association of Colleges and Universities
• Joseph B. Fuller, Professor of Management Practice, Harvard Business School

‍

‍

Michele Chang (0:00): Welcome, everyone. Again, this is the Entry-Level Hiring in the AI Era webinar today that we are hosting with Strada Institute for the Future of Work. My name is Michele Chang and I'm the chief strategy officer at Strada Education Foundation. I'm very excited to be with you all today to talk about this exciting topic.

Before we get to dive into the incredible data we have lined up, we have a couple quick housekeeping items. You are all entering muted with your videos off, but we want this to be an active conversation. We have the chat box available for conversation, and if you have any questions for our panelists as we move along in the webinar, you can use the Q&A box that you can find in your menu as well.

Today, we'll be exploring a vital crossroad: How artificial intelligence is redefining what employers look for in early-career talent, and the implications for higher education and employers.

We have a great lineup today of speakers to share their thoughts, as well as the research that we have found. First, we have Andrew Hanson, senior director of employer alignment at Strada Education Foundation. Andrew is going to share with us the findings and analysis from our report that he has so diligently done, and we are so excited to hear from him.

Following Andrew's presentation, we will be joined by three premier voices in the intersection of talent, education, and business – Alex Alonso, who is chief knowledge officer at the Society for Human Resource Management, also known as SHRM; Ashley Finley, vice president for research and the senior advisor to the president at American Association of Colleges and Universities (AAC&U); and Joe Fuller, the professor for management practice at Harvard Business School.

To get us grounded in the facts, I'm going to run and turn this microphone over to Andrew Hanson to walk through the research and the findings.

Andrew Hanson (1:49): Awesome. Thank you, Michele, and good afternoon, everyone. I'm sure many of us have been reading the headlines over the past year or more, claiming that entry-level jobs are disappearing because of AI-based automation. Now, if that were true, it obviously would have huge implications for our collective efforts to connect education and opportunity, and to help more people get good first jobs and launch their careers.

So we wanted to hear directly from employers on this topic to get an in-depth understanding of what they're experiencing. So in March, we surveyed about 1,500 employees and asked them questions like, “Is AI increasing or decreasing entry-level hiring and how is it changing entry-level work, like the skills you value and the mix of jobs you offer?”

The results, as I'm sure you'll see, will paint a more hopeful picture of the future than what we're reading in the news. A few notes on the survey before we get into the results. Everything that we're looking at today refers strictly to entry-level hiring in the U.S. So if you hear me just say hiring, you're safe in assuming that I mean entry-level hiring.

The respondents include a mix of executives and senior talent leaders who had ultimate responsibility for hiring, most commonly CEOs, presidents, and general managers. It was for businesses and organizations who hire at the entry level. We had, as I said, nearly 1,500 employers in our sample, which is representative across industries, employer sizes, and regions. In the survey, we looked at four core topics or areas: the volume of entry-level hiring; how employers were using AI; the factors driving entry-level hiring decisions; and work-based learning.

So with that, let's dig into the results. What we first saw was that as I noted, in spite of the doom and gloom that we’re regularly reading about the threat of AI to jobs, employers do seem to have more positive sentiments about how it's affecting the volume of entry-level hires, at least so far.

We first asked employers directly how the use of AI tools affected the volume of entry-level hires in 2025, and the anticipated effect for 2026. As you can see, they were very positive, about four times more likely to say that AI helped them increase entry-level hiring in 2025, and three times more likely to anticipate that AI would help them increase hiring in 2026.

Similarly, we looked at what was driving these shifts. Companies expanding their entry-level hiring head count credited AI as a major tailwind for growth, while companies pulling back largely did not point to AI as the culprit. Instead, they pointed to classic macroeconomic factors like economic conditions and the overall quality of the candidate pool. We also asked employers about how they're using AI, and found that the vast majority, about 9 in 10 employers, are at least exploring AI, with 22 percent saying they had strategically integrated it, meaning that they had a clear, company-wide plan for using it across their teams and business units.

We wanted to see what separated the companies who use AI to grow from those using it to downsize. The most notable differentiator we found was how integrated AI was. Companies who are increasing head count were far more likely to have comprehensively integrated it across their business functions, as we see here, which is that the more employers that integrated AI, the more likely they were to say that it is increasing the number of entry-level opportunities at their firm, both for 2025 and looking ahead to 2026.

We then asked, “What kinds of jobs are you increasing and decreasing head count for?” And first, for employers who said AI was helping them increase head count, they pointed to a mix of tech and business roles. Nearly 60 percent of these employers said AI was helping them increase tech roles like data science, IT, software development, and cybersecurity. Well, roughly 40 percent said it was helping them boost business roles like marketing, retail, warehouse automation, and process management.

For employers who were decreasing hiring, we asked which roles they were decreasing, and they similarly pointed to a mix of business and tech roles, the most common of which were administrative roles, customer support, and data analytics. You'll probably notice a paradox here, which is that both of these lists contain tech and business roles. The takeaway is that AI isn't eliminating departments. It's rather … it's filtering out specific tasks, which is leading to an evolution in the kinds of roles in each of these departments.

So after jobs, we wanted to know about skills: How our entry-level skill demands evolving because of AI, and which skills are most important? For employers who at least explored using AI, we asked how it's changing the nature of entry-level work.

As we look at this slide, I'll note three takeaways from the results. First is that AI is having a widespread effect on the tasks involved in entry-level work. We saw that only 20 percent of employers said it's led to no meaningful change, and less than 10 percent in the tech sector. The second is that across industry, AI is simultaneously reducing or substituting for routine and administrative tasks, but it is also increasing or augmenting the higher-order tasks that involve analysis and judgment. So many who claim that entry-level jobs are going away suggest that most of what's happening is substitution. But in reality, we're seeing both substitution and automation augmentation at similar levels.

The third is, as you might expect, this evolution of the tasks involved in entry-level roles is especially concentrated in the tech, manufacturing, and finance sectors. So on the whole, as we look at this, we see that AI is indeed affecting entry-level work, in essence, moving in a way from the sort of “grunt work” toward the kind of work we historically associate with mid-level roles.

Now, looking specifically at employers who typically hire at the bachelor's degree level, we asked employers to rate seven skills based on how important they were for entry-level roles, as well as how recent graduates performed on them. On importance, as we look across the entry-level job market for college grads, we see that foundational abilities like critical thinking, communication, and collaboration were rated as the most important, while AI literacy was actually rated as the least important.

Also, notably, employers rated performance lower than importance on every single skill except for AI literacy. And so, you know, at least in the eyes of employers, their perspective seems to be that the incoming talent has the tech skills down, but it's those foundational abilities where the gap continues to exist. Interestingly, as we look at the hierarchy of skills here, it is essentially unchanged compared to, say, a decade ago.

We then asked employers how they make hiring decisions and which factors are more or less important. What we see is that across the entire entry-level job market, work experience was rated as far and away the most important factor. The bachelor's degree had the second-highest rating of critical importance, and more than 70 percent of employers said that a referral from a trusted source was at least somewhat important.

For employers who typically hire college grads, we ran a little experiment. We handed them seven different candidate profiles featuring a mix of internships, regular jobs, project-based work experiences, elite academic credentials, and campus leadership experiences. We asked them to stack-rank their choices in terms of who they would most and least prefer. As we look at the results, two major takeaways stood out. First is that work experience beat out elite academic credentials across the board. The second was the ultimate differentiator really wasn't prestige, it was relevance: Was the work experience in a related role in my industry? The particular mode of how candidates acquired that work experience mattered less.

Then the last topic we looked at was how employers evaluated college graduates with different majors. We first asked, again, employers who typically hire college graduates to rate the performance of their college grad hires across the three different skill domains that you see here – critical thinking, communication, and collaboration – from 1 to 5, ranging from poor to excellent. The main takeaway that we saw was that no single major stood out as the best preparation for each of these core skills.

Math-intensive STEM majors were rated the highest on critical thinking. Communications majors were rated the highest on, you guessed it, communication, and arts and humanities majors, along with communications majors, were rated the highest on collaboration.

Finally, we asked employers who typically hire college grads to rank majors based on how difficult they were to hire to get a sense of the relative supply and demand of these majors.

As you see, math-intensive STEM and math-intensive business majors were the most difficult to hire, while employers had relatively less difficulty hiring arts, humanities, social sciences, and communications majors. We did ask this question both for generalists and specialist roles, but didn't see any interesting variation in the results.

So I’m very excited for the upcoming panel discussion.

I'll just offer a few concluding remarks. Obviously, we're still in the early chapters of AI, but my own bias is that the collective evidence warrants a tone of cautious optimism about the future. We're seeing that the first rung on the career ladder doesn't seem to be disappearing, but it is being rebuilt at a higher level. And, you know, as I noted, much of the so-routine, repetitive grunt work that allowed many of us to get our footing in the workplace is being automated, at least to some extent.

But it is also being replaced by these higher-order tasks that require these core critical thinking/communication/collaboration skills. So the dual challenge ahead for our collective work, as I see it, is one, how do we ensure that the pathways to entry-level jobs keep pace with that higher baseline? And two, how do we help more learners get access to the kinds of work experience that employers value?

So with that, thank you. And I will pass the microphone back to Michele.

Michele Chang (13:08): Great. Thank you, Andrew. Before you go, Andrew, we just had one question that maybe you can help clarify from the research that you had done. There was a question of whether entry level means GED, high school diploma required, or just some level of college degree or postsecondary training.

Andrew Hanson (13:23): Great question. The definition we gave is three or fewer years of relevant work experience across the entry-level job market. But as I noted or sort of flagged for you across the analysis, we were able to break the data down based on employers who typically hire college graduates, as well as those who typically hire at sort of middle-education levels, like, for example, associate degree or non-degree credentials, as well as jobs that require a high school education.

Michele Chang (13:49): Great. Thank you so much, Andrew. And now again, I'd like to welcome our panelists – Joe Fuller, Alex Alonso, and Ashley Finley. Thank you so much for being with us today. I'm really excited about this conversation. As Andrew shared, we had so many interesting insights from this report, but we know there's a lot more than just what a report will say.

So we're excited to get the perspective from these three exceptional individuals. So first, the first thing we want to dive into is how much AI is helping employers increase entry-level opportunities, which, as Andrew highlighted, was a surprising finding from our report. We'd like to really dig into what did you all make of that result, that employers were more likely to say AI was helping them increase, rather than decrease, entry-level roles, and that companies with higher levels of AI integration were the ones that were more likely to use it as a way to expand entry-level opportunities.

I'd actually like to start with Professor Fuller to say his thoughts on this, as I know he's done a lot of work on looking at the future of work.

Joseph Fuller (14:51): Well, Michelle, thank you for having me and pleasure to be with your audience. I think the study is a pleasant surprise, but I think the results are what we'd say are “early-innings” results. Similarly, a paper which tracks with your findings by Kathryn Bonney and her team at the U.S. Census Bureau, called “The Microstructure of AI Diffusion,” a fancy-sounding title, showed some pretty similar results, but it also showed that adoption of AI and the expansion of employment related to AI was in large companies – 18 percent of enterprises, but 32 percent of head count, often supported by technology vendors, generally limited to one of three functions – marketing/sales, business development/strategy, and IT.

The reason I'm saying this is an early indication, but one we shouldn't be too quick to assume is an indication of long term, as large companies focus in a few functions which have demonstrated business cases for the application of AI, does not speak to the sweep of a general-purpose technology. There is some other data that's a little more unsettling, such as the unemployment rate for master's degree, recent master's degrees, graduates being 50 percent higher than the three-year trend line, the unemployment rate for 20- to 24-year-old college graduates being 15 percent higher than the trend line.

So a lot of different signals here. Early innings, really interesting, provocative data. I'm sure we are all hopeful that the trend will continue. But if you look at predictions of what was happening with things like the deployment of mobile technology and internet, there are a lot of false negatives and false positives early in the process.

Michele Chang (17:12): Great. Thank you, Joe. Alex, would you like to chime in there as well?

Alex Alonso (17:16): Yes, I just wanted to kind of share a little bit about some of the research that we've seen recently, which sort of validates both what Andrew and Joe are talking about in terms of these being positive, yet also sort of early-innings results.

When I think specifically about the data that we capture on a weekly basis, we capture roughly data from over 25,000 employers on a recurring basis, specifically looking at what it is that they're doing in response to the adoption of AI and mass adoption of AI.

To Joe's point, for the larger employers, they're able to bring back and actually reshuffle work the way that they think they should do that. To give you context, what we're seeing is on a professional level or industry level, that means really looking at how work gets done and what is qualified as entry-level work. Then beyond that, going into what qualifies as not entry-level work.

Because of that, what we're actually running into is a space where we see from our own data roughly 57 percent of employers, and these are the HR leaders of those employers, saying we are now redefining what the work means to identify what entry-level work looks like.

As you think specifically about what we're seeing, we're actually looking at in some cases there are professions, whole professions, that are going through a great truncation kind of movement, which is we're redefining what the entry-level work looks like, but we're also redefining what the most strategic creative-level work looks like.

So that creates a scenario where, yes, we may be adding new definitions of what entry-level work looks like, but at the same time, the overall whole of that profession is actually shrinking, stagnating, or having a different kind of perspective. I'll give you an example, what we see more than anything else when thinking specifically about our labor economic view into this, we tie that to the behavioral work that we see with our employer population.

What we find more than anything else is, if you look at it today, there are 22.6 million jobs today that are not subject to displacement because of what we call AI. As we think about it since 2022, they're actually subject to automation. Andrew keeps using the term “grunt work.” In reality, I think of it coming from my working-class parents, I think about there's no such thing as “grunt work.”

But one of the things that stands out is we keep seeing this notion that work is actually much more highly automatable over and over again. If you were to think about it, that's people who have at least 50 percent of their jobs today that could be fully automated. When we think about that entry-level work, we should think broadly about what that falls into, because we're in this great churn that for some professions means full truncation and disappearing, but for other professions actually means a complete revolution in a different way.

So I raise that largely because there's great context that needs to be accounted for. But this is remarkably positive looking at this work. I do think there's an opportunity to look at that. By the way, I didn't say this up front, but just like Joe, I'm ecstatic to be here with you all. I'm grateful to be part of the panel.

Michele Chang (20:48): Great. Thank you, Alex. I think you are highlighting one of the things that we did want to dive into a bit more was about how employers are saying that AI is both helping them do what you might call some of the more routine groundwork, but also increasing tasks that involve higher-order skills that involve more judgment analysis.

I think you're highlighting great things that we're seeing that potentially is happening here is the redefinition of what entry level means and perhaps some movement to more mid-level type of roles.

Before we move on from that, Joe or Ashley, anything you'd like to add on that topic before we talk a bit more about what this means for credentials and skills and majors?

Joseph Fuller (21:26): Just a quick observation, Michele, which is we see in our data at Harvard and also in some recently published Anthropic data, that AI utilization really peaks at around the 70th percentile of income and then diminishes pretty quickly thereafter, which is an anomaly, because certainly in our analysis, if you put wage data as a log of wages and compare it with an anticipated penetration of AI against task, it's almost a slope of one up to like the 95th percentile.

You really expect those higher-wage workers to be using it in a higher rate than they are. I think what we're seeing is some demographic matching that companies are putting veteran employees with low technical aptitude but high contextual intelligence, and they're hiring some young talent to support it. Certainly, we're telling some of the companies we're working with directly on a consulting basis, think about building your teams as demographically diverse, not so much in terms of race and gender as it would have been five years ago.

But put that 30-year veteran with the second-year employee with the CS degree from Carnegie Mellon. You're going to put technical intelligence, contextual intelligence, and you might get something pretty exciting.

Michele Chang (22:53): It's a really interesting insight. Ashley, did you want to jump in there?

Ashley Finley (22:57): I will just say something quickly and just to join in. And again, also say, I'm so excited to be part of this conversation. Delighted to be with you all.

Just as we're on this particular topic, I think it's worth mentioning, my background as a sociologist is in work and occupations. The push-pull of labor, of occupations, since the Industrial Revolution is the push-pull of automation and specialization. So it is worth underscoring as we have each new data point.

I realized as we were having this conversation the visual in my mind is that connect the dots where like the really good ones, where you don't actually know exactly what the picture is going to be until you connect all the dots. It's like each new data point gets us a little closer to seeing what this picture looks like.

But this is the push-pull of the history of work, of labor, is to automate. We are yet again in a really exquisite tension point of what it means to live with that level of automation and how exactly it will disrupt work. It will recede and we will figure it out. That's what it's reminding me of.

Michele Chang (24:12): Thanks, Ashley. Just a quick reminder for everyone on the webinar as well. We do have the chat function. I see a number of folks are getting in there and introducing themselves. But also we do have a Q&A function. If you have any questions for our panelists at the end of our panel discussion, we will take a few questions from the field.

The next one we wanted to dive into is, obviously, there's a lot of emphasis on AI skills today, but employers are continuing to value foundational abilities such as critical thinking, communication, and collaboration. As Andrew noted, the hierarchy of skills is basically the same as it was a decade ago.

Ashley, I'd love to pull you into this conversation and see how you have seen how educators are thinking about which skills employers are valuing, and how those efforts have been made to integrate AI skills into curriculum.

Ashley Finley (25:01): Again, I'm reflecting a lot on how we think about quantitative literacy when calculators are a fluid part of how we perform that function. Does it still matter if my kids are able to do long division, will they move on to other things? I do think, one of the … well, resonating with the idea that we are still talking about a similar set of skills.

In fact, one of the things that I think, if anything, we're missing on that list or that I would encourage people listening to be thinking about, is the way in which things like curiosity, resistant resilience, persistence, that growth mindset, right? This notion of how dispositions and mindsets that come in to work with those skills and how they are consistently applied.

We know from our own employer research that those are not the fluffy things on the sides. Those are absolutely as critical and essential as collaboration, critical thinking, communication skills. The ways in which we think about augmentation and the ways in which we think about how AI is augmenting and enhancing those skills, it seems quite resonant with me, and I hope is great news for the kinds of things that we've been doing for a long time.

And the curriculum. I think it is great news for the kinds of things we're still doing in the curriculum. What will be increasingly important is that students understand they are requiring those skills. We still need to have the same conversations about students being empowered to recognize how and where they have built those skills, and are able to communicate those to employers.

Michele Chang (26:47): Thank you. Ashley. Alex, you want to jump in there?

Alex Alonso (26:49): Yes, sorry. You can see the excited person in me. This is actually something that I think is more often than not overlooked. Which is it is the same skills that are needed. We have such a desire to want to define this construct of what is AI fluency and the skill set around it that we don't actually think about what are the core inner cognitive processes and abilities that we really are looking at in a different use case, more often than not, not actually building new fluencies.

What I find particularly interesting is, if you were to ask most employers today, just same sample, same group of employers, we know that there are roughly 9 percent of employers who have a definition of what they consider to be AI literacy. That means 9 out of every 10 do not have a context, an actual functional definition of what they're looking for when they're defining an AI literacy kind of definition. What I find particularly galling about that is we use that as a criterion for making decisions around what it is that we're looking for, for talent, how we will evaluate talent, and then beyond that, for any partnership.

When you look about public-private partnerships, when we look about educational institution partnerships, employers wanting and seeking talent, that brings together some sort of skill. But we can't even define the skill that we're looking for and what proficiency and/or performance looks like.

So we at SHRM, as an example, one of the things that we've worked on over the last quarter is actually defining what AI literacy looks like, but more importantly, doing so in a way that is actually user-friendly, but comes with a look at what are those power skills, those skills that are the ones we've always known about, those cognitive processes, which ones are the ones that actually lead to this notion of AI literacy?

But more importantly, what is productivity? We mean incremental performance above and beyond what you get with the standard cognitive abilities and/or personality traits, other skills, abilities and traits that we think of when we think about the traditional model of KSAOs.

My inner IO psychologist is just coming out now, and what we find is so true to what Ashley was just saying, the levers for driving what our higher levels of proficiency and productivity when it comes to AI, what predicts further are things like curiosity, are things like not your computational skill, but your ability to say, “This is the problem I'd like to tackle,” and actually being able to communicate that in a way that can pull context and deliver context back and forth when thinking about that. So those are the levers that actually drive that.

We see that those relationships actually have real collinearity or can correlate with what we're looking for, as opposed to thinking specifically about what we see traditionally, which are the things that you would think of across any kind of tech revolution.

Joseph Fuller (29:54): Just to build on what Alex was saying, Michele, for a second, I think it's borne out in a very good article late last year in the Harvard Business Review on the phenomena of what's called work slop, which is AI-generated content, which at first blush seems to be substantial, but with any kind of further inquiry proves out to be essentially shallow, indefensible conventional wisdom. Workers are reporting that 40 percent of them had received some substantial work slop in the last month, and it accounted for 15 percent of the work product they had received.

They were tending to get it from peers, but they're also getting it occasionally from their supervisors. So you have to have imagination, the ability to take a risk, be a resilient raise with someone. “This work doesn't make sense to me. Are you sure?” All the types of social foundational skills we've been talking to in the last 10 minutes gain a premium in this world because this is a general-purpose technology.

It's going to be integrated. It's better thought of as a decision to change the way you structure and manage work than acquire a technology. It's going to become so integral to processes, involve other humans, that all the foundational skills are going to be escalated in importance for that reason, as well as the fact that some of the technical jobs in each guild, in each position, will get automated or augmented away, and by act of transitivity, what's left as a greater ratio of the skills you're seeking are those skills that automation was not even close to replicating, the foundation skills.

Alex Alonso (31:46): Just to layer on a little bit, Michele, I know you're trying to move to the next, but work slop is a genius term. But the latest trend is actually what we're seeing, that 82 percent of workers and supervisors agree they are reporting it's not just work slop, it's work slop lasagna. That is where you pile work slop on top of other work slop on top of more work slop. Like any good lasagna, it tastes good going in, but it makes you sluggish on the way out. The productivity actually slows you way, way down. What's fascinating about that is not only do they report actually seeing it, another 70 percent say they engage in it, meaning that they actually practice it as part of their performance. That is a dangerous phenomenon.

Michele Chang (32:33): Definitely don't want that kind of lasagna. So we're going to have to be careful. Thank you for adding that.

Well, we did want to also touch a bit on work-based learning. A big theme of the results that we saw was that work experience was a key differentiator on who's more likely to get hired, so that it's really hard when you think about jobs at the entry level, it's a bit counterintuitive to think about work experience for someone that you're hiring at the entry level.

So I would love to put it to this panel. How should we think about better integrating work experience into postsecondary education? And maybe Ashley, you could get us started given your role in postsecondary education?

Ashley Finley (33:12): Yes, happy to, particularly given the way in which we talk about high-impact practices, which are really effective and engaging educational experiences that we've long talked about in higher education, things like learning communities, capstone courses, first-year seminars.

There's no reason that we shouldn't be thinking about what it means to have a work-integrated or work-based experience in any one of what we might call “curricular anchors” as really these kind of touchpoints that we often are seeding throughout the curriculum as points of opportunity and deep engagement among students.

What we fail to do in those spaces is really define what work-based means, what does work-integrated mean. One example is Guttman Community College in New York, Bunker Hill Community Colleges Massachusetts do a program called Ethnographies of Work. That's a required first-year or entry-level experience in which students are going into a workplace, any workplace, and effectively observing, evaluating, and analyzing the culture of the workspace. That's a really important and interesting work-based experience that can happen early on that's deeply connected within the curriculum and embedded within disciplines.

I think part of our bigger issue is really thinking about what does a viable work-based experience or work-integrated experience look like. I think a lot of times you're just asking students to reflect on what are the skills involved. What is my relationship to these skills? Who am I? What is my temperament in relationship to this? What problem is being solved?

So this is really important, critical and fundamental things that can be done all along the way, and how we think about what those work-based variances are beyond the internship. That's really what we're talking about here is, “What's the difference?”

If we're only talking about internships, I think we have a sense of what that means. But unless you're that, you're probably one of those campuses that are trying to scale internships, really challenging to do. I appreciate Strada’s work, quite frankly, on what work-based means, what it means to have paid experiences for that, and just how we think much more robustly beyond the internship is the only place, the only way we're thinking about that.

Michele Chang (35:42): Thanks, Ashley. Yes, it's actually highlighted for those who haven't followed, Strada has done quite a bit of work on work-based learning and really brought in the definition of what it is and quite beyond internship. So certainly urge you to take a look at some of the reports that we've done there. We've had a recent report on who's actually getting paid with work-based learning experiences. I definitely encourage you all to take a look at that.

Alex, did you want to chime in there?

Alex Alonso (36:04): Yes, I would just say, you know, we have actually engaged and this is actually in partnership with groups like Strada and other groups like Stand Together and Charles Koch Foundation. One of the things that … and Walmart Foundation, forgive me … one of the things that we've looked at specifically is what it is that recruiters and/or HR professionals, executives, and hiring managers see, what it is that they are looking and thinking about, what it is that drives their perception of work experience.

What is it that you would specifically look at and say, “This is the work, what I mean by work experience”? Because not all work experiences are equal, right? Adding work experience elements to the education is always a critical component. But one of the things that we want to step further … because it was fascinating, you see work experience across the board, it didn't matter which sample we were looking at, said this is No. 1 and this is what I'm looking for when thinking about the most important thing. I want to know where they worked. I want to know the kind of work that they did, and I want to understand the kind of output that they're capable of.

One of the things that we actually went back to, and this is the same samples roughly, I want to say, almost 2 1/2 years later, one of the things we asked them is, “What would you, in lieu of work experience, what would you accept in an era of AI as potential substitutes for that, that would draw your eye?” The No. 1 thing that CHROs or hiring managers or recruiters or anything like that said, was, “Well, I'd love to be able to talk to any digital asset that this person created from the use of AI. So for instance, if they created a digital twin, if they created an assistant, if they created an agent, let me test that out. Let me blow it up and let me see what I can learn from it.”

No. 2 is show me how you interact with your own AI, not just the prompting, but show me the depth of skill and thinking that you throw at this as a substitute for, “OK, so you don't have work experience, but prove to me that you have those entry-level skills that I can throw you at bigger problems down the line and that prompting and people's prompt approaches.”

Really, their prompting was the best way to demonstrate that for a given series of professions and/or industries. To me, that was particularly telling, because if you had asked 2 1/2 years ago, not a single one of them would have said anything about AI. But all of them are adapting, knowing that they have to adapt when thinking about the talent that comes forth.

That in and of itself not only points to the transformation that we're talking about, but to Ashley's point, starts to look at how we actually build out alternatives that really make the most sense, especially in the preparation process, but beyond that.

Michele Chang (38:47): Joe, did you want to chime in?

‍Joseph Fuller (38:50): Yes, I think there's a hopeful signal in this discussion and we can build on the early hope that the excellent study speaks to. For years, employers have been very reluctant to invest in creating lots of new opportunities for work-based learning. Even when you show them great results from other companies and other programs, they want often some near-guarantee that engaging in accepting work-based learners, whether it's co-op, earlier internships, apprenticeship programs, federally registered or not, we can put all that hair splitting aside.

Now with Workforce Pell coming up, it's very reluctant. How do I know that's going to work out? This could be expensive for me. I could get sued. I could get unionized. The more they require some proof of capacity to do a job and interest in and durability in doing a job. So they don't run the risk of someone who's actually studied something but actually he's never done it and doesn't know that they want to stay doing it, the more they're going to be attracted to, “I want to rent to own. I want to see this person in my workplace, see the foundation skills, see the drive, see the aptitude, see the grit.”

I think this may finally get us on a path where more employers are saying, “I can't trust credentials.” Ninety percent of recruiters and 50 percent of hiring managers report seeing applications AI supported that have false claims, false credentials, false representations of AI knowledge.

More reporting. A third of hiring managers and 40 percent of recruiters are reporting more in-person interviewing, now with more-focused, targeted questions. The next logical step is, “Let me, let's say, hire the college sophomore; heaven forbid, a college sophomore with an internship. If they're good that summer, I'll hire them the next summer. If they're good that summer, I'm going to give them a job offer, which includes me paying for their senior year in college if they come back or with a demand note behind it. It's not going to just be here, take a scholarship.”

But people are going to have to go there with a velocity of technology. With the rate the AI LLMs are changing, your 201 level course in computer science you took as a sophomore will be n-minus-five technology when you're hired after your senior year. I can't rely on the report card any longer.

Michele Chang (41:54): It's a tough, changing world. Joe, something you said made me think of something else we had discussed as we were preparing for this panel. Obviously, we're talking a lot about how AI is impacting the type of skills employers are looking for, but we also cannot overlook the fact that a lot of HR managers and hiring managers are utilizing AI as they are looking for entry-level candidates as well.

Just curious if you all might be able to opine a little bit on what you're seeing from that perspective and how that's also changing and shifting hiring at the entry level.

Alex Alonso (42:28): I'll jump in first, just because I probably spend the most time with hiring managers and HR professionals. One of the things that we're seeing specifically in the recruitment process is a tension that exists predominantly in the use of AI.

We have a series of regulatory bodies who are pushing for this notion of if there is a final employment decision that needs to be made about a human, it should involve a human set of judgment in it, and specifically not just use AI in any way, shape, or form. What that has done is had trickle-down effects in terms of the way organizations are actually starting to use AI relative to what the original HR roadmap may have looked like prior to that regulatory approach.

Predominantly, it was in the world of recruitment and talent acquisition and more importantly, talent assessment. How do I go about actually judging talent and/or evaluating talent and their ability to actually do the performance and the criteria that we needed? That was the first boom that we saw. That has actually led to some recession because of the regulatory nature of what we're seeing, however, predominantly in New York City, Colorado, and California.

But what we're seeing is now a movement where we're seeing a push toward looking at a couple other things. In other words, we're seeing groups that are focused primarily on embedding AI in the recruitment process to unlock and provide sources of new talent and thinking about how you might bring in talent that you may never have considered.

So much so that we're actually seeing examples of this pop up. For instance, one of the things that I've actually seen as an example is the notion of taking heavy mobile equipment mechanics from a large employer, a federal employer, that looks specifically at how they look at their skills and taking them because they have all been apprentices and/or mentors in some way, shape, or form, and porting them to educator roles as a primary example, entry-level educator roles to say, “OK, this is how we do that.”

It came because of an AI-based solution that actually said, when you look at the career map of these or the profession map of these two kinds of roles, there is a lot more between their DNA than you might actually think. That's an example of where AI is being applied to see solutions that we don't normally see. The sort of the same way that you think about pharma and new drugs and new medications in that regard.

The other thing that I think is particularly telling, though, is we're seeing it actually move away from what is that talent identification sphere and now moving into the talent nurturing sphere, and really doing more to help organizations and employers think about how they nurture both potential candidates, how they nurture a pipeline of talent coming their way, and what it is that they're doing now.

The natural input there is, it is an actual perfect benefit, if you were to think about it, a marriage, if you will, between the pipeline of professionals, future professionals that are coming in as entry-level professionals, and employers, large employers in particular, thinking about how their AI chooses to nurture that pipeline. Before you know about it, you actually then begin to create a symbiotic relationship that says, if I'm an educational institution or a pillar of academic integrity and rigor and development, one of the things that I might look at is, how do I take that workforce development that I am putting forth in terms of the way we nurture these power skills and make it so that it actually stands out in a variety of these pipeline management and/or pipeline nurturing kind of AI that exists out there.

The employer community is signaling those and saying, “Please come back and give us that talent, and we will make these available to you so you know what we're looking for and how we want even that pipeline to develop. We will feed them back to you and make sure that you can redevelop them in ways that we need.”

Those are just two examples of where we're seeing real AI application and changing the behavior of employers. I'd love to see how that actually leads to changes in the partnership and the communication and throughput that happens between educational institutions and employers.

Michele Chang (46:43): Thanks, Alex. Joe and Ashley, did you want to jump in there? We were going to move to some questions from the field, but happy to give you a few seconds if you have something you'd like to chime in on.

Ashley Finley (46:55): I’ll jump in quickly, Michele, just because I actually saw one of the questions in the chat was around equity and what are the implications for equity and inclusion as we think about some of this?

Alex just laid out a lovely point around how AI is used to actually maybe discover latent skills in jobs and in a person's work history that we might miss. It reminds me a lot, this is going to be a little bit of a throwback for some of you, but reminds me of the early, early days of the Music Genome Project, where Pandora and Spotify … Pandora was one of the first to do it, but would put together a playlist. You're like, “No, I wouldn't, I never would have. I loved this song, and I never thought to put it with this song.” It's because it's analyzing the latent characteristics of the music, and that is what we are going to be able to do. This is actually one of the Wild West of what we're talking about.

No question. Still so much yet to be seen. One of the ways I get really excited about what I hope we will see is the way in which we do discover talent in ways that never had a chance, in ways that we … the ways in which we honor pathways in the arts and humanities, the ways we have a much better understanding and visibility of what are the skills being built across a number of different pathways, the way we might be able to honor work study in richer and deeper ways than we have before because we're able to analyze using AI. I just so appreciate that point by Alex and want to double down on that.

Michele Chang (48:29): Great, thank you. All right, we have a number of questions from the participants on the webinar. We're going to try to get through as many as we can. Then we'll come back to the panel for some closing remarks. I'd like to also welcome back Andrew Hanson, as there are some questions that he may be able to help provide some color on.

We've got a couple of questions curious on how our survey findings impact the hiring of graduates who studied computer science or IT and what obviously, in the media, we've heard a lot of attention around what's happening to these graduates in particular. So I think the questions around … any thoughts around those type of graduates and what their career prospects may look like.

Andrew Hanson (49:07): Really great question, and something that I think has been top of mind for many of us at Strada. Obviously, we saw over the past decade or so just an explosion in the number of students studying computer science and then a more recent decline or perhaps the regression in the past couple of years, like thinking about this signal and perhaps over correcting, let's say.

My own bias is that the death of computer science has been greatly exaggerated. I think that certainly the survey, if you look specifically at the mix of jobs that are included, the evolution of tech roles as opposed to the complete automation of it as like what we see in that data. But a number of other interesting data points from studies that have come out in the field, such as the continued growth of software developers, the employment of software developers over the past couple of years.

I think it is something that we're paying attention to. I think we don't want to jump to conclusions too quickly before this sort of plays itself out. But I think there's still promise there.

It wasn't a major topic of what we looked at, but the other I should have mentioned, the other relevant result was the majors. Still, math-intensive STEM majors continue to be very difficult to hire for in the minds of employers and evaluate as well. Certainly, as we look at the earnings of folks who study computer science, they continue to be in the top three. So that's why I share.

Michele Chang (50:49): Thank you, Andrew. Another question we have here is what recommendations or thoughts do you all have on how AI is impacting recent graduates, both at the associate level but also in the skilled trades?

Alex Alonso (51:10): I would actually … go ahead, Joe.

Joseph Fuller (51:13): I'm just going to say in a model that I'll be publishing soon with extensive research, what you see is that really no job is unaffected. The degree of effect in skilled trade is less. But it often affects things like storing measurements, scheduling workers, ordering inventory, submitting proposals.

Interestingly, right now there's a company that sells AI-equipped drones to roofers, and they fly over your roof and do all the measurements. So the roofers look for a roof that's failing and send the homeowner, “I took pictures of your roof. It's failing. And here's my bid because I've measured it using an AI-supported camera.” So there are elements of skill trades that are hard.

It's not quite the dirty dark dangerous we use in manufacturing and mining and ag, but AI doesn't matter what career. It can do a lot of the dull, dispiriting, and deterministic work that people would love to off-load so they can do the main tasks that they enjoy.

Alex Alonso (52:46): The only thing I'd add to that is I see prime examples of that happening all over the place, where you're bringing people into a profession or changing materially, how that skilled trade may actually shift.

The example I think of is I sit on the National Council for Credentialing Crane Operators, and they credential close to 600,000 crane operators every year. One of the things that stands out is their future is actually one that's driven more toward drone technology and AI-based technology for spatial reasoning. As we see that, there are a whole new set of skills, opportunities, and professionals that are more likely to be able to move into that trade and see opportunities to have greater transportability into the skill set that they're bringing that you wouldn't have seen 20 years ago. So that's an example where all of a sudden you're seeing real potential.

But to Joe's point, it's in the targeted approaches, unless you're seeing a wholesale shift in the technology that they support.

Ashley Finley (53:51): Michele, can I add one quick point here that I don't think has been lifted up? I keep being mindful of the time, but we're talking a lot about how AI changes, like very specific tasks, you know, task-oriented skills, which makes a ton of sense.

But I really appreciate SHRM's work on civility and in particular, SHRM's work on developing a civility index and calculating the amount of lost production that is of consequence that is borne out by having incivility in the workplace. We know from our own AAC&U employer research that fully two-thirds of employers say that students’ ability to reason through and manage civil discourse or disagreement, constructive dialog across disagreement, is very useful in the workplace.

I'm thinking about colleagues that I know that have rerun an email through AI and said, “Make this nicer, tone this down a little bit.” So I just wanted to add that in that there are these human elements of how we think about teamwork, the fundamentals of collaboration, and how invaluable it might be to take a beat and run something through AI that's going to help with those kinds of skills, too.

Michele Chang (55:06): Great. Thank you, Ashley. As we near the top of the hour, I just wanted to do one more quick round for everyone on the panel and just get how are you feeling about the future of work and where our education and our workforce systems are going, and where career pathways need to go from here? Joe, do you want to start us off?

Joseph Fuller (55:27): Oh, sure. I'm probably more bearish in the intermediate term than many, certainly many of the people that study this in the academy. The reason for that is twofold, threefold maybe. One is that the rate of progression in the last year in the LLMs is astonishing. And 80 percent of the venture capital that's been invested in the last two years in the United States is in AI native offers.

So you're going to get a tremendous continuation of the innovation explosion. Every time you get a job that is somehow insulated or protected from AI's penetration, that's a market signal that tells investors and entrepreneurs and inventors, if you can solve this problem and unlock a lot of extra productivity with the innovations in a model, you're going to be rich and famous.

Finally, I just want to refer to something Alex touched on. If you look at the way AI is being deployed in China, it's much more integrated with the forms of automation. And we are post-industrial society largely. It may in fact help reshoring when we can have more or less lights out of plants here based on AI with automation.

But we haven't seen that type of adoption yet in the U.S., and that's going to lead to another wave of disruption. I think we'll get a lot of productivity improvements and quality of human life and quality of follow-on invention, like in healthcare. But I think it's going to be a rocky 10 to 20 years in terms of labor market rules and organizations going to have to challenge changing the way they approach both creating skilled people and deploying them as they try to get their heads around how do you actually manage in the era of a true general-purpose technology.

Michele Chang (57:36): Thank you, Joe. We're at the top of the hour, but Alex, to just give you a quick couple of seconds to maybe have some closing comments.

‍Alex Alonso (57:43): Well, I will say, you know, the big fear is that Skynet is here. I joked with my fellow panelists yesterday about Skynet being here.

I don't actually believe that. I think that we are, to Joe's point, in an era where we are actually seeing the greatest return on investment. Specifically, workforce investment comes from a marriage of AI plus HI, as we call it at SHRM, and I'd argue that the better-suited economies will be the ones that actually leverage that and build systems to make as much AI plus HI as possible.

Ashley Finley (58:16): I would just add for the higher ed folks in the audience, really trying to figure this out, getting their AI policy sorted and dealing with students’ anxiety around AI. Stay curious. We're still connecting the dots on this, but there's lots of really good evidence to suggest we're still on the right path. The path we've been on is still the right path, but we need to work harder at helping students to communicate their skills well.

Michele Chang (58:42): Excellent. Thank you all today for being with us, both the participants and our panelists. A massive thank you to Alex, Ashley, Joe, and Andrew for their time and insights today.

We do want to note that there is a link in the chat to download the report that we've been mentioning, so you can reference that. But thank you all for all the work that you all do to connect education and opportunity.

Have a great afternoon.