WGU Labs begins experiments in embedding human connection into AI-first systems
With the rapid spread of Gen AI, we’re entering a time when human connection is becoming more valuable and more vulnerable.
That’s a reality that postsecondary institutions need to confront head-on to both remain relevant and to grow their impact on students’ long-term outcomes.
Social capital has always been a core ingredient to student success. Even before the widespread adoption of Gen AI, an estimated half of internships and jobs came through personal connections. Students with access to mentors are significantly more likely to graduate feeling prepared for life after school. Young people with an adult encouraging them to pursue their goals are more than twice as likely as those without to have a promising future. And for students from low-income households, connections to individuals from high-income households are the leading predictor of economic mobility.
Put simply, who students know matters.
Yet despite their immense value, relationships are not actually core to the business model of most traditional higher education institutions. Glossy brochures and websites may market students learning in community with faculty and one another, but colleges and universities don’t earn revenue based on how connected their students are. And few institutions are held accountable for graduates’ outcomes—including whether students leave campus with the networks they need to get the jobs they want. Case in point: a mere 9% of graduates say their alumni network was helpful in the job market.
Leaving relationships to chance is no longer a smart or sustainable strategy. Other forces are looming that stand to simultaneously exact costs on students’ networks while also boosting the value that connections will carry in the labor market.
To begin, the rise of Gen AI may spell the end of entry-level work as we know it; the supply of entry-level roles stands to shrink dramatically in some industries in the coming years. And in many industries, “entry-level” roles are increasingly requiring multiple years of work experience.
At the same time, hiring processes themselves have become infiltrated with AI, ironically leading to a surge in referral-based hiring.
And finally, while connection commands a greater value, it’s also getting displaced. Students are increasingly leaning directly on AI—rather than humans—for help, advice, and support. While that may have short-term upsides, it threatens to gradually weaken the very networks they’ll need to build and advance in their careers.
How can postsecondary institutions meet these compounding connection challenges at the intersection of AI, the labor market, and student support?
The opportunity isn’t to eschew AI, but to make it an engine for connection. Institutions need to start building and scaling AI tools and learning experiences that are fundamentally prosocial.
Prosocial designs put tech in service of human connection, not in place of it. And prosocial tools can make it easier for students to find, reach out to, and rely on others.
Last year, we launched a partnership with WGU Labs to test new approaches to doing just that.
Exploring prosocial possibilities with WGU Labs
At first blush, WGU might seem like a strange champion of social connection; it is, after all, the largest asynchronous online learning provider in the US. That means its students primarily engage in coursework at their own pace, and on their own.
But WGU is also one of the most innovation-forward institutions in the country, and its research and development arm, WGU Labs, was created specifically to pursue early-stage innovations that could improve the postsecondary learning field—for both WGU and a wide range of other 2- and 4-year college partners.
Like many organizations, WGU Labs’ interest in prioritizing social connection grew hand in hand with the immense opportunities that Gen AI presents to postsecondary providers, like spawning new AI-native course design tools and building personalized student support chatbots.
The team saw that scaling these new tools could be a double-edged sword. AI could radically democratize efficiency and access, but, implemented poorly, it could further isolate students from crucial human relationships and atrophy their skills and confidence to ask others for help and advice.
Instead, WGU Labs wanted to pilot designs that offered students the best of both worlds: the affordances, flexibility, and adaptability of high-tech tools, with the enduring value of high-touch human relationships.
Capturing new data on relationships and help-seeking
Before launching into building prosocial tools, we worked with the WGU Labs researchers to design a survey for current WGU students. The survey captured baseline data on the state of students’ social capital. Their findings revealed stark realities and some troubling gaps:
- Nearly 1 in 5 WGU students (19%) reported knowing zero people in their desired career field. Over half reported knowing 3 or fewer.
- WGU students from lower-income and first-generation backgrounds report consistently weaker professional networks, up to 10 percentage points behind their more affluent, continuing-generation peers.
- 77% of students said they felt they belonged at WGU, but only 28% felt connected to other students.
- 76% said they do ask for help, but 81% said that all else equal, they prefer to handle things alone.
- Only 53% of students said they had talked to anybody in the field they’re interested in about advancing their career in the past month.
- Only 36% said they felt comfortable reaching out to WGU alumni.
As WGU Labs researcher Stephanie Reeves summarized, “these results point to a structural disconnect: students are motivated and clear on their goals, but they’re not being supported in building the networks that help translate a degree into opportunity.”
For Reeves and her colleagues, the data reinforced their commitment to exploring fundamentally new ways of wiring social connection into their work. “These findings underscore the need to redesign digital learning environments with human connection at the center,” she wrote. “That means going beyond discussion boards and affinity groups to build systems that intentionally foster relationships between students, alumni, and professionals in the field.”
To start imagining what those systems could look like, we worked with the WGU Labs team to build on what we’ve learned over the last decade of studying innovative strategies for building students’ social capital.
Building new solutions anchored in promising practices
Early on in our work on social capital development in schools, we realized that a deliberate strategy to tackle social capital gaps must address two related but distinct dimensions: students’ access to social capital and their skills to build and mobilize it.
In other words, do students have access to high-quality relationships—with peers, educators, industry professionals, and community members—that can lend them resources like information, advice, and opportunities? And also, are students developing the mindsets, skill sets, and confidence to tap into their existing networks and build new ones?
Within these domains, the WGU Labs team determined what indicators of both skills and access were most meaningful to their work—and best aligned to where the survey data had revealed room for improvement. They chose to focus their initial round of pilots on three core indicators:
- Increased access to peer connections,
- increased access to professional networks, and
- increased helpseeking confidence
As WGU Labs researchers and designers began to consider ways to design toward those outcomes, there were a number of evidence-based and practical guidelines we urged them to consider:
- Beware of point solutions that shortcut the social capital development process: While in any pilot project it can be tempting to test out small dosage interventions, narrow point solutions are unlikely to unlock the transformative potential of social capital, especially if they skip the key steps of defining and developing social capital literacy and understanding existing membership within and across networks. Forthcoming research from researchers Jonathan Zaff and Astraea Ausberger suggests that social capital development follows a process: Defining social capital & developing social capital literacy; then visualizing networks (through maps and metaphors, ideally in a cohort setting); then experiencing networks (in guided activities and low-stakes practices); then building networks on their own out in the wild.
This process runs counter to many approaches to “teach networking” or “host networking events” that operate in a vacuum, without a coherent developmental arc. If interventions or tools encourage students to do one or a few of these things in isolation—such as building a LinkedIn profile—they’re unlikely to have a strong effect on a student’s overall social capital development.
- Focus on building conversations and shared experiences over engineering relationships: Trying to build relationships without understanding how to foster conversations could amount to lots of matching and meetings for students without those efforts adding up to meaningful connections, much less building their confidence to go out and build networks themselves. If pilots aren’t producing higher rates of human conversation, and if those conversations aren’t deemed helpful, it’s unlikely that institutions will successfully promote a future in which AI-powered connections and human connections are striking a healthy balance.
- Imagine a world beyond 1:1 support: For many education providers, 1:1 support is considered the gold standard. But while individual attention is valuable, it can ironically miss opportunities to develop social capital across the student population. Youth development research has long shown that webs of support are critical to resilience and success. The prevailing assumption is that a single caring connection or mentor can help students in every circumstance, yet in reality, we often need to turn to different people when we face different challenges and opportunities. A successful social capital strategy may mark a departure from 1:1 support structures toward group structures that are more resilient and diverse.
- Connect students with near peers for know-how and senior professionals for hiring: Efforts to increase students’ access to professional networks should reflect the diversity of types of connections that expand opportunity. It’s long been shown in the research that different types of relationships can confer different benefits. For example, strong ties tend to lend care and support, while you’re more likely to find new information and opportunities through weak-tie networks (the so-called “strength of weak ties”). Distinctions like this can be important to overlay on measurement strategies, where it can be tempting to default to “the stronger the relationship, the better” thinking.
More recently, research has revealed that different types of relationships may have different impacts on students’ career journeys. The Search Institute found that among a group of education-to-career providers, near peers proved the best source of information and support for students. This is consistent with prior research on near peers as credible messengers and as possessing more relevant, real-time knowledge of labor market realities.
Still, there are limits to what near peers can do. Research from Basta, focused specifically on first-gen college students’ job hunting, revealed that mid- and senior-level professionals provided the most significant resources to students, well ahead of both community connections and near peers. This is consistent with findings from Raj Chetty and his team at Opportunity Insights, which suggest that access to cross-class connections is one of the leading predictors of economic mobility.
Social capital pilots
With these core principles in mind, the team at WGU Labs has launched a series of pilot projects to test how best to introduce social capital skills, practice, and connections into tools and online experiences.
First, they’ve begun piloting a credit-bearing social capital course that teaches students about the value of social capital and helps them reflect on the networks already surrounding them and how to mobilize them in support of their goals. The course includes an interactive relationship map, a series of reflections meant to ignite a growth mindset to network-building, and a series of formative assessments to gauge changes in student confidence over time.
Second, they’ve built a new tool designed to place students in cohorts with peers so they can support one another academically and professionally. The tool aims to match learners with their interests and backgrounds, and facilitate resource sharing across student cohorts. It’s designed to complement traditional mentor-driven supports, rather than replace them.
Third, as part of a larger effort to build personalized student support tools, they’ve embedded nudges for students using career support chatbots to reach out to people in their networks for help and career advice. The model draws on WGU Labs’ Professional Capital Framework and includes helping students set smart goals and align social capital development opportunities with those goals, as well as interview simulations to build skills and confidence.
Across all of these efforts, the team will be collecting data on how students’ networks and confidence are or aren’t developing over time.
Seeding what’s next
These pilots are now underway, with data coming in later this summer and fall. Like any tech-forward experiment, the team is seeking to understand both what works (i.e., what drives meaningful increases in social capital ) and which scaffolding and user experience designs drive engagement.
These efforts are not only important advances in the field of teaching social capital (embedding prosocial principles into courses and support structures, and scaling students’ professional networks), they’re also poised to inform broader conversations about the future of postsecondary learning in an era where institutions often reach for technology to build efficiencies rather than connections.
The technology and research exist to flip that script and build toward a fundamentally more social future of learning.
