from
Over the winter holiday, on a family trip to North Carolina’s Outer Banks and the neo-ancient splendors of Charleston and Savannah, we ended up in Atlanta. Walking around the cold, dark downtown on the final night, we stumbled across a restaurant chain my kids had never encountered: Ted’s Montana Grill. Zev asked why the logo was a bison. So I told them the tale of Robert Edward Turner III, one of the most remarkable men of our time.
I talked about how, by befriending and beguiling new cable operators across the country, Ted Turner took an Atlanta UHF TV station and turned it into “Superstation” TBS. I talked about Ted’s vision for CNN and how he singlehandedly (and unfortunately, but inevitably) established the 24x7 news cycle. And in the 90s, his marriage to Jane Fonda; Jane moved into Ted’s apartment at CNN and every night Ted would walk his movie-star-aerobic-queen wife home through the sports marketing department. Eventually I wound my way to bison: Ted selling his company to Time Warner and becoming a billionaire; getting duped by AOL, reducing his wealth from tens of billions to billions; but how he still had enough to buy a lot of land in places like Montana and repopulated bison across the plains; and how he was so successful that he had to come up with a strategy to cull his growing herd. Hence Ted’s Montana Grill, offering a more environmentally friendly, lower fat alternative to beef.
Zev: So he’s both a hero and a villain to bison?
Me:
Exactly.
My favorite Ted Turner stories don’t involve business or bison but rather sailing. Throughout the 1970s, while he was building Turner Broadcasting, Ted was one of the world’s top sailors and captured the 1977 America’s Cup. After his boat’s mast broke during one America’s Cup trial, he was interviewed by a TV reporter:
Reporter:
What did you think when the mast broke?
Ted:
I thought, there goes the mast.
And following the infamous 1979 Fastnet race off the coast of the UK, when an unprecedented summer storm sunk five boats and killed 15 yachtsmen (but not Ted, the race winner), the Mouth of the South was at it again:
Reporter:
How can we avoid tragedies like this in the future?
Ted:
You ought to be thankful there are storms like that, or you’d all be speaking Spanish. It was a storm like that that sank the Spanish Armada.
There was no instruction manual for Fastnet 1979. Even if there had been, Turner wouldn't have read it. Famously impatient, the man probably never followed an instruction manual in his life. Come to think of it, I don’t know anyone who reads instruction manuals. And the good news is we don’t have to anymore.
***
As workflows shift from physical to digital and organizations invest billions to gain promised efficiencies, learning to use and take advantage of increasingly complex software becomes a choke point. Fortunately, there’s an answer – one with ramifications for the future of work and education.
The first digital adoption platforms (DAPs) were released just over a decade ago to reduce employee frustration, help companies realize a return on software investments (because there’s little hope of a return if no one uses it, or if no one uses it correctly), and – thankfully – replace instruction manuals. DAPs sit on top of complex business software to guide, cajole, and teach. DAPs like WalkMe, Whatfix, and Apty engage employees from within applications to deliver competence and hopefully mastery.
The DAP concept isn’t new. Those of us old enough to remember Microsoft Office’s Clippy from the late 90s have nightmarish visions of a personified paperclip with bug eyes and a penchant for providing irrelevant information at inappropriate times. Pretty much everyone hated Clippy, perhaps because the “Office assistant” was designed for first-time users. Most of us turned Clippy off and eventually Microsoft did as well; by 2001, Clippy was no longer an Office default setting.
Although the idea of in-app guidance remains the same, DAPs are different. They include recording technology and authoring tools allowing employers to demonstrate workflows and create step-by-step walkthroughs. And critically, as new features are deployed, DAPs alert employees and upskill them in-app without the need for scheduled training.
Matt Stewart, founder and CEO of RiseNow, a procurement and supply chain digital solutions company, believes DAPs should be considered in every rollout of a new procurement platform. According to Stewart:
Complex systems like source-to-pay platforms can be overwhelming for most employees, especially if they’re not using them daily. DAPs are becoming essential to drive adoption, foster behavioral change, and achieve ROI. There are many DAP use cases in source-to-pay alone: supplier onboarding, approvals, requisitioning, non-catalog purchases, and invoicing.
But the biggest difference between Clippy and DAPs involves learning. Although as anthropomorphized as a paperclip could be, Clippy didn’t want to learn anything about us. (That may be the ultimate source of the vitriol.) In contrast, DAPs adapt with each click. If we’re making a mistake over and over, when we’re about to reach that point in the workflow, the DAP interjects with a helpful hint. But once we do it right, DAP stays silent. Like Big Brother or the Chinese Communist Party, DAP is always watching (but in a good way). If it spies a workflow that could be automated, it asks whether we’d like to do that, then shows us how. And if it sees employees aren’t changing or optimizing usage despite tactful suggestions, it sounds the alarm.
There are two takeaways from the fact that the market for DAPs is already $1B. First, just how important and complex software has become. Second, having rendered instruction manuals obsolete, DAPs may do the same to classroom instruction.
***
Last year I wrote about how, despite the hurry and hoopla around generative AI, there was little evidence it was about to change the classroom experience in high school and higher education. I cited one observer who found AI tutoring “unconvincing,” stating that “AI tutoring today seems to consist of breaking down problems into component parts and explaining the components. This is no doubt helpful, but it is not tutoring in the true sense of the word.” And another who believes we’ll have flying cars before we have effective AI teachers. My point was that AI was already having a salutary effect on administrative efficiencies in education, and that’s where we should focus our attention.
But I’ve come around after seeing what DAPs are already doing. DAPs are brimming with AI, providing guidance when users are struggling and anticipating mistakes before they’re made. Now imagine similar functionality across not only procurement, sales, and finance, but also compliance, customer service, and creative fields like communications, design, and product development. If we’re working with software, intelligent DAPs built into that software will ensure we’re learning while working and getting better at our jobs.
Learning to use business software is different from learning to think. But if the software is sufficiently complex, how different is it really? Mastering complex business software requires a wide range of cognitive and durable skills like critical thinking and problem solving. It also necessitates understanding the industry in question as well as fundamental business functions and processes. So what if AI’s primary impact on education isn’t in the classroom, but rather shifting the locus of learning to outside the classroom? What if AI flips the default from the classroom – at least from high school and beyond – to project-based learning, experiential learning, learning by doing, and – best of all – earning and learning?
Instead of sitting in a classroom listening to a teacher, high school and college students could be assigned real work and learn from that work. Students could be matched with employers or specific projects provided by or derived from employers, then do the work on the same software used in the enterprise. As AI-powered DAPs become increasingly powerful, they have the potential to transform real or simulated work into educational best practice for students only a few years away from seeking full-time employment.
If DAPs take us in this direction, four implications come to mind:
1. Clearer path from education to work
It’s hard for educators to know exactly what new talent employers need, especially if employers don’t know themselves. But here’s a surefire way to address the problem: have students work on the software platforms used in the jobs they want to get. Proficiency and problem solving on the same software will simultaneously narrow the
skills gap and
experience gap. And if the software can’t be identified, it’s probably not a job worth training for or a program worth offering.
Working with complex software platforms will also expose students to AI agents. While most high school and college students are already taking advantage of ChatGPT or Claude for studying, research, and writing, they’re unlikely to be exposed to agentic AI in an academic setting. In contrast, where the goal is to get a job done – rather than learning for learning’s sake – AI agents will proliferate. And within a few years, that’s where most graduates will need to be to land a good first job: prior experience working alongside autonomous AI “colleagues.”
2. Who teaches?
Today’s high school teachers are typically education majors and teachers first, subject matter experts later (hopefully). An AI-driven work-centered education model upends this structure, foregrounding projects and work over teaching methodologies and learning styles. Instead of grading multiple choice quizzes and five-page essays, high school teachers will need sufficient subject matter expertise to understand, assign, and evaluate real work. This ought to mean more diverse pathways into teaching and greater fluidity between teaching and the rest of America’s workforce – where subject matter expertise matters a great deal – allowing more experts to teach and more teachers to find gainful employment outside education. High school teachers won’t disappear, but they’ll be different.
As for college, while faculty are subject matter efforts by definition, for many there’s a chasm between their expertise and the world of paid employment (unless of course such employment involves perpetuating the subject matter by teaching future students at a college or university). So it could become paramount for professors to have worked outside higher education and done meaningful thinking on connections between skills derived from their subject and how those skills can be applied in the world of work. If they can’t do that, they shouldn’t be taking tens of thousands of tuition dollars from students whose primary objective is good first job and career launch.
3. Leveling the playing field
It's also good news for boys. As I wrote about
last fall, boys have fallen behind at every level of our education system, largely due to insufficient motivation and aspiration, driven by the fact that current classroom-based learning makes biologically unrealistic demands of future orientation and impulse control. Teenage boys and young men will do better in high school and college with work-based learning and more immediate and relevant feedback.
4. Better retention
Finally, a work-based model for secondary and postsecondary education will be good for learning. Learning is only retained if it’s used. MIT researchers from the Center for Brains, Minds & Machines have concluded there’s
no evidence that anything significant and unused is remembered beyond two years. If we care about retention, more learning should occur in the work software and modality where students find themselves a few years later. It may seem like a small step, but it will be a giant leap forward from textbook exercises and essay prompts.
***
Are you tired of prognosticating puff pieces with dull descriptions of AI-powered learning such as “personalized learning pathways” which “adapt in real time”? Pundits continue to blandly accept individualized learning will continue to happen in and around classrooms. To my mind, this assumption is as lost at sea as boats in Fastnet 1979.
Ted Turner went to college at Brown but was kicked out for having a female student in his room. He never earned a degree, instead learning through work and developing deep interests not only in the environment and bison, but also global affairs: Ted’s Goodwill Games played a role in ending the Cold War and his Nuclear Threat Initiative has helped secure weapons of mass destruction for the past quarter century. Ted learned by doing.
In the digital century Ted helped usher in, it’s fair to ask whether – after elementary and middle school – there will be any learning without doing. Anything else could become an anachronism, or worse, widely acknowledged as a waste of time. And although we all have Fonda classroom memories, that’s something impatient Ted wouldn’t have sailed, sat, or stood for.