Building A Digital Ladder

Many college faculty leave a lasting impact, but I’m only aware of one with a national holiday. Although no state is giving people the day off (yet), November 9 has been designated Carl Sagan Day by fans of the Cornell astronomer, author, and TV personality. Every November 9, we’re encouraged to read his books, look at the stars, celebrate the beauty and wonder of the cosmos, and “ wear a turtleneck sweater with a brown jacket.”

My first encounter with Sagan’s questionable turtleneck + brown jacket combo was Cosmos, the most successful show in PBS history. As I’d never heard someone talk in paragraphs before, let alone eloquent passages linking Ancient Greeks, evolution, and interstellar travel – come to think of it, I’d never heard anyone who didn’t end a sentence with eh? – I was hooked. For my 10th birthday, my dad obliged by donating to PBS so I could get the Cosmos book and Sagan-brown tote bag.

Perhaps the most memorable Cosmos segment is Sagan’s presentation of the Drake equation where he estimates the likelihood of detecting signals from intelligent life elsewhere in the Milky Way. The equation comprises functions for number of stars, stars with planets that could support life, planets where intelligent life actually arises, intelligent life that evolves a technical communicative civilization, and the final function – f(l) – “the fraction of a planet’s lifetime graced by technical civilization.” f(l) was denoted by a mushroom cloud, conveying – jarringly to this 10-year-old – that productive technologies invariably arise in tandem with destructive technologies, and intelligent life may not have – in the words of Sagan (or, more accurately, the Johnny Carson parody of Sagan) – “billions and billions” of years. Thousands might be a stretch.


One thing we know for sure about f(l) is that it’s digital; f(l) will be written in 1s and 0s. Although one day chips may no longer be made of silicon and it will be commonplace for digital signals to converse with neurons, f(l) is clearly and irrevocably digital.

Since the dissolution of the Soviet Union three decades ago, despite many missteps, America has retained its position as the world’s foremost power principally due to its digital leadership and economic gains therefrom. Nearly all the most important digital milestones have taken place here and, not coincidentally, nearly all the biggest digital companies are here, making money off software and data. (China now has digital giants, but only one – Bytedance – spans beyond the Great Firewall.) And at least until recently, global talent flooded into our companies and colleges.

So far so good. But two obvious risks might make f(l) shorter than anyone would like. The first: after nearly 80 years of Pax Americana, large scale conflict with an authoritarian regime. The likelihood of this appears to be a function of a second risk: American unity and stability – an authoritarian regime is much less likely to test the resolve of a united, stable America (or, conversely, only an unstable America would risk initiating, escalating, or responding in a planet-jeopardizing manner).

America has always had crackpots. One (now obvious) downside of “productive” digital technology is that today’s crackpots have amplifiers and echo chambers. But the reason rage and resentment have become so pervasive is that while nearly all Americans are consumers of digital technology, a sizable proportion are shut out from the production side. Which means feeling shut out from the dynamic digital economy and meaningful economic advancement – particularly problematic as the verboten fruits of the digital economy are on full display for all digital consumers. Which means not recognizing the future because they don’t see themselves in it. Which means being as susceptible to grievance and lies as, say, Germans were in the 1920s and 30s, in the wake of national humiliation and hyperinflation.

Whether f(l) is drastically shortened because a single deluded American unleashes devastation in response to the FBI searching Mar-a-Lago (or a single deluded American tries to hide nuclear secrets in Mar-a-Lago), or whether f(l) is the product of a small coterie of officials unleashing slow-motion destruction (e.g., blocking action to combat climate change), the best response to a political party threatening violence, rooting for civil war, and willing to trade away rule of law, democracy, and a livable climate so long as they can “own the libs” is to build a digital ladder so their intended audience doesn’t feel left behind. Once good middle-class digital jobs are within reach of every American – or at least visible, with reasonable hope of attaining them – Harry Potter’s Chamber of Crackpot Echoes will be much smaller.


How to build a digital ladder? It starts by rethinking our approach to education. Digital skills are often interpreted as coding. But coding is a small part of what’s missing. Digital transformation has progressed to the point that much of the necessary coding to make the economy whirl has already been done. So a digital ladder must equip job seekers with what can be best characterized as platform skills. For example, CRM ( Salesforce), HR ( Workday), finance ( NetSuite), sales and marketing ( Hubspot), customer service ( Zendesk), software development ( Atlassian), cloud computing ( AWS), data warehouse ( Snowflake), and digital transformation itself ( ServiceNow). Moreover, each industry has its own platforms like hospitals ( Epic), insurance (Applied Epic), home care ( WellSky), construction ( Procore), pharma ( Veeva), legal ( Clio), lease administration ( Lucernex), sports (Thapos).

Although these platforms require a good deal of functional or industry knowledge, they don’t take years to learn. As a result, they ought to be within reach of tens of millions of frustrated Fox News viewers. But mastery is unlikely to arrive in a classroom. These skills are much harder to learn in class than by doing. Mastery comes via practice: learning-by-doing.

Susan Wright, director of Talent for Good, a Salesforce apprenticeship program, explains:

In Salesforce, automations reduce clicks, make things more efficient, and improve data. We want automations. But you can’t really teach this in a classroom. To successfully build automations, you must do the technical work of building the Flow, then test, document, and roll out to users.

Data Management
This is another Salesforce item that can seem logical in a classroom, but once you get into it, things get complex. You load the data and encounter a slew of errors. Error handling cannot be taught in advance. You must hit errors and interpret and deal with them. For example, what if you import data and forget to turn off automations? Then you have a slew of records being updated and triggering automations such as creating other records, editing records, or sending emails? You can teach the steps to execute. But experience is the only way to learn how to complete the activity successfully and with validity.

Problem solving is clearly necessary for Salesforce success. But it’s a long way from sufficient. John Stallings, Talent for Good’s program manager, concludes:

There are numerous great resources available to learn Flow or Data Management, but until you get hands-on with these concepts, it’s incredibly difficult to connect the dots. Students need to have the opportunity to learn about these topics, then take it a step further and put things into practice with hands-on exercises and capstone projects.


While digital transformation has placed a huge premium on learning by doing, learning-by-doing is something colleges have a hard time doing. There are three primary reasons for this:

1. Faculty

Learning-by-doing doesn’t come naturally to faculty who earned their credentials in classrooms, not by doing. Few Ph.D programs involve much in the way of applied learning. So learning-by-doing is not the way they were taught, and therefore not held up as an exemplar of academic excellence. In addition, relatively few faculty have experience as practitioners. As currently constituted, the academic treadmill leaves little opportunity to jump off and do something else for a few years. Many faculty have never worked in the private sector (at least not since frontline jobs undertaken in high school and college summers). Even fewer have used Salesforce, Workday, or Hubspot. And even if colleges relaxed Ph.D requirements, as digital transformation continues to accelerate, the most sought-after experts in the highest value sectors would probably continue doing, not teaching.

2. Hard to Do

At a recent meeting on the skills gap (we all get together occasionally, mainly to laugh about how we’ll never be out of work), I raised the heightened importance of learning-by-doing and challenged the group to come up with solutions for schools. One participant suggested that if colleges gave freshmen laptops, the first order of business could be to take them apart and put them back together. That’s a great idea, I said. What happens in week two? Or for that matter, what happens each week of all 40 courses constituting a bachelor’s degree (approximately 500 course weeks)? Are 500 relevant learning-by-doing exercises within reach? Likely for computer science, engineering and courses that already have labs. Languages and performing arts are also plausible. But humanities and social sciences are a challenge. For the most popular college courses, faculty would be hard-pressed to come up with one learning-by-doing activity every week for every class.

3. Absent Employers

Neither faculty nor difficulty would be a major impediment to widescale adoption of learning-by-doing if millions of employers were knocking on campus gates begging faculty to incorporate relevant real work into coursework. But employers are only conspicuous on campus by their absence. They simply have better things to do.


Building a digital ladder means prioritizing new learning-by-doing models. And the best digital ladder is getting paid for learning-by-doing. In stark contrast to unpaid internships, getting paid to learn is the only educational model where opportunity isn’t dependent on one’s background or resources – the only educational model where underrepresented minority, first-generation, LGBTQ, rural, religious, or Republican students aren’t at a disadvantage.

In opposition to college’s “train and pray” model, this digital ladder is apprenticeship – a job where workers are hired first then paid while they train on the requisite platform skills. It turns out everyone loves apprenticeships. In a recent survey, 92% of Americans had a favorable view.

Apprenticeship is the rarest of commodities in modern America: something we all agree on politically. The New York Times reports “the desire to expand apprenticeship reflects a rare area of bipartisan agreement.” Perhaps rational Republicans recognize apprenticeship could help awaken the party from its fever dream. And apprenticeship is perfect politics for a Democratic party struggling to win back the support of working class voters who now overwhelmingly view it as the party of a liberal, college-educated elite. As Eric Hoffer recognized in The True Believer: Thoughts on the Nature of Mass Movements, “mass movements can rise and spread without belief in a god, but never without belief in a devil.” Historians are telling President Biden that America is facing something akin to pre-WWII proto-fascist movements. As Republicans like “Little Marco” Rubio attack these same historians as “elitists” and “snobs,” it’s clear that for Trump’s would-be stormtroopers, the devil is the college-educated elite.

One reason we’re struggling is that there’s a huge disconnect between what we believe in and what we invest in. Funding for apprenticeship is a fraction of funding for “train and pray.” By my calculations, federal, state and local governments continue to invest over $400 billion each year in the analog ladder known as college while total spending on apprenticeship is under $400 million (1/1,000). Comparing a single apprentice to her college-attending friend, support received by the apprentice is about 2% of what we spend on the college student. There’s simply no way many of the apprenticeship activities employers are currently obliged to fund in their entirety – delivering related technical instruction, mentoring, even hiring unproductive workers – don’t have a materially higher return on investment for students and society than much of what’s spent on accredited postsecondary institutions each year.

America needs a digital ladder for our digital economy even more than Carl Sagan needed a better outfit. So it’s urgent that we foreground learning-by-doing and background classrooms, at least until we reach a more balanced approach to funding and status. We must do this quickly. Because the stakes couldn’t be higher. And we don’t have billions and billions of chances to get this right.