Higher education was born, in places like Oxford and Cambridge, as a tutorial system. But over the centuries, higher education moved further and further away from that model through waves of expansion and democratization—reaching a point where the 500-person lecture hall came to be synonymous with college.
Similarly, apprenticeship was our first system for developing new workers—a multi-year process of learning alongside an expert tradesperson. But as the world went through waves of industrialization and the corporation rose, the training of workers took on the same mass quality as other areas of production. These changes occurred not because mass education was most effective, but because individualized instruction couldn’t scale to educate millions.
Both sectors have spent the past 20 years investing—with mixed results—in software solutions to change that calculus. But now, with AI we can see the outlines of a world where it’s actually possible to have hyper-personalized education, learning and career development at scale.
Generative AI enables what’s known in the world of tech investing as “service-as-software.” It’s a term of art for a world where services that could previously only be done by humans can be replaced partially or fully by technology. For a sector so dependent on services, this shift opens up new solutions, new business models, and new markets focused on unlocking human potential. Here’s what that means in real terms:
New Solutions
Until now, technology-enabled companies focused on human potential have faced major limitations. They lacked the human-like reasoning, conversational interfaces, and context needed to support the complex processes around education and learning. Just look at some of the core jobs around developing human potential:
- Course building, teaching and tutoring
- Academic advising
- Career coaching
- Skills assessment
- Career pathway design
- Skills-based recruiting & hiring
- Corporate training
- Manager development and coaching
And the list goes on. To date, software has only been able to assist with a small portion of the work typically done by people in those roles, and human capital development has remained slow and costly.
But with AI, suddenly we can create custom training courses, tailored to the needs of specific businesses or kinds of learners, for pennies on the dollar. We can do qualitative assessment at scale, and we can provide on-going feedback rather than waiting for a high-stakes exam (at school) or a performance management cycle (at work). We can offer truly personalized career coaching. We can create avatars that allow learners to practice interacting with patients and customers. And this is just the start.
Of course as we build out these solutions, we will need to balance technology’s growing role in providing information and support to learners and workers with the knowledge that prosocial behaviors and human relationships with mentors/teachers are still central to helping people flourish. More on that in a future post.
New Business Models
That in turn, enables a completely different business model for edtech and HR tech. Many classic ed-tech funds are collapsing or re-branding, as most companies in the sector have seen either poor outcomes—because they don’t provide the necessary human wraparounds—or poor returns—because they do.
AI can change those prospects. By dramatically reducing the cost of human-like support and communication, AI shifts the potential profit margins in edtech and HR services companies from the 30% that is typical in service businesses to the 75%-plus margins seen in software companies. This shift to service-as-software creates working capital to invest in building quality experiences and getting those products in the hands of more people.
New Markets
By lowering delivery costs, the model also expands the market—opening it to lower-income learners and workers who have a limited capacity to pay and traditionally have been more expensive to serve. Similarly, this new model opens the market to small and mid-sized businesses because the cost to personalize learning and development solutions for each SBM owner and their team has come down dramatically with AI.
Better Outcomes
Strategically, this means we now have a new toolkit for solving long-standing problems in learning and talent development, and for the new ones that will arise with AI. We can provide the kinds of personalized experiences and wrap-around support that research has long shown is best for learners, workers, and anyone looking to grow. AI also allows a shift to “always on” support that simply isn’t possible with human advisors and coaches.
In other words, AI presents an enormous opportunity to drive better learning and career outcomes even as we drive down cost.
To be sure, some readers will think this is an overly-optimistic view of what’s possible in impacting incredibly complex challenges. All the more so given the turbulence among top tech companies and the still very-real limitations of today’s AI.
Historically, people overestimate technological change in the short-run but vastly underestimate it in the long-run. The vision laid out here is what’s possible over the next decade if we start building for it now. And we do have to build for it intentionally. The problems traditionally disadvantaged workers and learners face can’t be solved by technology operating in a vacuum. These are systems and culture problems, and it will take real effort to ensure that the most marginalized populations actually benefit from AI.
A win for workers, learners, and society would be a full-circle back to the tried and true learning that we know works from apprenticeships and the tutorial system — but now with the scale needed for modern demand.