This Is Not the Industrial Revolution

And it’s not just pompous jackasses like the not-so-strawman above. I have read 116 policy papers on the future of work... They are all wrong.

This Is Not the Industrial Revolution

“This is just like the Industrial Revolution; AI will create more jobs than it destroys,” he informs me after my keynote on the future of work. “I don’t know why you’re so worried about it. You should read the World Economic Forum report on how AI will empower everyone. Just like all of the weavers that founded the fashion industry, AI will create super doctors by taking away all of the busywork and leaving only the fun stuff. We’re already teaching coal miners how to program. Now we need schools to teach every kid how to program...and STEM and AI, so we can send them all to university. Programming is the future of work! And social skills! AI can’t do social, so we can train everyone else’s kids to be elder care workers. We don’t even need universities anymore; we can just send them all to trade school to get upskilled. I didn’t attend your keynote but the problem with it is that ATMs actually created teller jobs. The gig and sharing economies give workers all the power to do what they want to do, but we might have to pay some people to do nothing. That’s what’s so great about AI: it always makes people more creative. It’s just like the Industrial Revolution.”

I somehow find myself face-to-face with these Rubenesque cartoons of late-stage male CEO-hood on a semi-regular basis. They’re always desperate to mansplain my own research to me. Many of the conversations start with a non-apology about not even having attended my talk, but then follow a rather vague tour through an echochamber of thought leadership pieces that they’ve skimmed, all of which supposedly prove that I’m wrong. I’m told repeatedly that AI will create more jobs than it destroys (something that I have never disputed). And always the mantra: “It’s just like the Industrial Revolution.”

I have read 116 policy papers on the future of work... They are all wrong.

And it’s not just pompous jackasses like the not-so-strawman above. I have read 116 policy papers on the future of work. IEEE, McKinsey, the White House, the UN, the UK, the World Economic Forum–everyone has a policy paper on the future of work. They are all wrong1.

I recognize how ludicrous it is to suggest that the armies of brilliant, informed authors behind these policy papers are all wrong. In defense of my arrogance, I first offer the virtual cut-and-paste quality of so many of these “insights”, all launched off the back of the original WEF “Fourth Industrial Revolution” concept. But much more meaningfully, they all seem to be answering the wrong questions.

Is today really like the Industrial Revolution? The answer is nuanced and complex (which never makes anyone happy) but essentially, no. Will AI create more jobs than it destroys? I don’t know definitively, but it seems very likely that it will create many jobs.

Rather than simply speculating on AI job creation, we should be asking an entirely different set of questions. Who will be qualified for these jobs? Did the rising automation during the Industrial Revolution directly increase or drive creativity, or is this just lazy myth making? Do we, in fact, misunderstand what the Industrial Revolution was2? And perhaps the most important question: where do creative people actually come from?

...routine labor, even the most complex and highly-educated, loses its economic value.

These questions have been the focus of my research for the last 15 years, and some of the answers and insights we’ll explore in this episode point to a very different future and past than today’s lazy myths suggest. We’ll look at how human capital is becoming a toxic asset as routine labor, even the most complex and highly-educated, loses its economic value. At the same time, we’ll examine the rise, both today and throughout the Industrial Revolution, of what economists call non-routine cognitive labor–what I call the creative economy: any job that explores the unknown.

These lazy myths sell us “inevitable” futures from Star Trek’s techno-utopia to Elysium’s dystopic oligarchy. In reality, we have choices but we must be willing to think about this moment in history in its full, messy3 complexity.

Take for example the following thought experiment. You are the CEO of a multinational company with 100,000 employees. Rate all of those jobs on a scale from "lowest" to "highest" skill4. Now consider a near future in which AI and automation have disrupted the bottom 80% of those jobs by skill-level. Those 80,000 jobs are not needed anymore, and those lower-skilled employees are staring at pink slips. But just as with the Industrial Revolution, automation, in this case in the form of artificial intelligence, has created an equal number of high-skilled jobs. So you have 100,000 employees and 100,000 great jobs–or maybe even more. This is wonderful. Problem solved, right?

But wait, now your company now needs five times as many high-skill employees. AI hasn't created any new lower-skill jobs because if they fall below the skill threshold then those jobs are in turn automated as well5. So ask yourself these questions: will many, if any, of those lower-skilled employees be qualified to fill these new top-20% roles in your company, even with reskilling?

Take a step back. Today, how easy is it to recruit for and fill those top-20% positions that already exist in your company6? How would that change if you have five, ten, twenty times as many “top jobs” to fill?

And what if we’re not talking about the top-20% but the top-1%? Will productivity boosts from AI lift your entire labor force into these elite roles? Do you truly believe you can retrain even a minority of your workforce to fill those new jobs?

I believe that we can7, but it isn’t going to be through reskilling or the gig economy. It won’t be because we’ve given everyone a university degree or taught them all to program. This is not the Industrial Revolution. In order to secure a robot-proof future for our children and our economy, we must stop pretending that it is.

Testing Lazy Myths

“Now we need schools to teach every kid how to program...and STEM and AI, so we can send them all to university. Programming is the future of work!”
– Pompous Mansplainer

Back in the day, every self-appointed hero with venture funding was launching an EdTech company that was going to save the world. I did them one better, not only gracing education with my genius but also rescuing industry from the intractable problem of biased hiring8. As I was building education companies (as well as continuing to dabble in my theoretical neuroscience research), I also explored what makes a great employee as Chief Scientist at Gild. Our tagline was, “We take bias out of the hiring process.” (Our other tagline: “We bring meritocracy back to tech hiring”. Unfortunately, research shows that the very use of the word “meritocracy” actually makes bias worse. So much for the self-appointed heroes.)

Nearly every major policy paper, and the wannabee thought leaders that quote them, says that university enrollment and programming skills are the winning combination for the next Industrial Revolution. My analysis of 11 million professional programmers at Gild completely disagreed. (And as we’ll see soon enough, it’s not even clear that the vast majority of code will even be written by humans in ten years.)

Our dumb little AI stumbled into the simple truth that it isn’t simply that you knows how to program but what you do with it that matters.

Gild built a database of 122 million people, pulling together hundreds of thousands of data points from more than a hundred different websites. From this painfully jumbled mass of information, we built targeted machine learning models to illuminate the qualities that predicted a great employee. Contrary to standard hiring practice, we found that the skills that people claimed on resumes, LinkedIn, and elsewhere were shockingly poor predictors of their performance on the job. The claim that a candidate knew the popular programming language “Python” on their resume (and often even proof of the language’s use on sites like Github) didn’t predict the actual quality of the code that they wrote. Similarly, many developers claim that they know “C/C++”, a pair of foundational programming languages. Our models actually learned that this resume claim was a negative predictor of their ability to code in either of those languages. For most serious C programmers it was akin saying, “I know Spanish/Portuguese!” Our dumb little AI stumbled into the simple truth that it isn’t simply that you knows how to program but what you do with it that matters.

Even a candidate’s university degree told us surprisingly little once we knew other things about them9. A bachelor’s of Computer Science from Stanford, which our own algorithm rated as the best computer science school in the world, was only a relatively modest (linear) predictor of the quality of code written by software developers. In other words, you don’t write better code…because Stanford.

By comparison, our model of an applicant’s motivation swamped all of these traditional hiring measures. Yes, motivation is hard to measure, but AI came to my rescue10. I built targeted machine learning models that allowed me to illuminate some of these deeper qualities about applicants11. Perhaps more surprising than motivation, social skills were just as strongly predictive of job performance for software developers as for sales people. Glib social skills might be less common in engineering than in sales, but you write code for people and with people. Our ability to understand others helps transform routine work into something creative. (It’s interesting to note that social cues and upvotes on Github were very much not predictive of the quality of code someone wrote12. Who could have guessed that social media and social skills are not the same thing?)

Most fascinating of all was that this emerging list of the best hiring predictors was eerily familiar to me. They were the same predictive factors that were emerging in my education research. Across wildly different populations and goals–life-outcomes of little kids, graduation rates of university students, and work performance of job applicants–traditional skills and knowledges weren’t first-order causes of success. The future was written in deeper, latent qualities about the individuals themselves.

Students and employees weren’t successful because they knew how to program or factorize polynomials, or because they attended a particular school; they were successful because of qualities like self-efficacy, endogenous motivation, working memory, and dozens of other constructs. What is amazing is that these constructs are as predictive for engineering as sales, for CEOs as athletes, and for adults as children.

By exposing what is not predictive of life outcomes, my work at Gild debunked dominant policy responses to automation: send everyone to university, reskill the workforce, and teach everyone to code. All three of these solutions assume that humans are endlessly fungible widgets whose economic value is largely defined in terms of a list of skills and knowledges. If the economy needs new skills, reskill. If it needs more advanced skills, go to college. This view assumes jobs are fundamentally routine labor. The only split is between routine cognitive labor requiring advanced education and non-cognitive “unskilled” labor. But another class of work dominates economic outcomes today, and probably has throughout all history. Many economists call it non-routine cognitive labor; I prefer to call it the creative economy.

...the core question isn’t how many jobs AI helps create, but rather who will be qualified to fill them.

The laziest myth of all is that the productivity gains of the Industrial Revolution gave birth to the creative economy. Many would have us believe that, nearly overnight, village weavers were transformed into the multi-trillion dollar textile and fashion industries, lifting people into creative labor for generations. This myth tells us not to worry about continued automation by artificial intelligence because automation always creates more, better jobs than it destroys13. But the core question isn’t how many jobs AI helps create, but rather who will be qualified to fill them. The slyest misdirection in this lazy myth suggests that those new and creative jobs of the past went to those whose previous work was automated. But was that ever true?

The collected works of Prof. Schlashenschlessen

“This is just like the Industrial Revolution; just like all of the weavers that founded the fashion industry.”
– Pompous Mansplainer

Annotated Journal. Jahosafats Goosepimpel. c. 1804

...was clearly Formica rubra. It was possessed of feelers four and unequal, exactly as the Vicar had described. I noted the singular articulations, placed at the tip of the lip, which is cylindrical and nearly membranaceus. T’was far in the woods outside the Farm a pleasant summer’s stroll – I was happy to do the Vicar’s work but arrived so late at home my backside red and numb from the treatment of Father’s switch.

My Brothers complain as they have for years that the Vicar treats me as though an apprentice – maths, readings, the natural world (soft work, they say, as those who cannot read have no measure of the mind’s labors). But they could do it too if they were willing to endure Father’s rough treatment for all missed chores.

I see not the bother that fewer have requested Mother’s weaving. With much new fabric on the market, Mother spends much time idle from weaving and no house chore goes undone.

In truth it was Turbish’s obligation to collect the specimens in the woods but as always the shiftless greenhorn failed his contract. I know the Vicar regrets accepting him as his apprentice… Indeed I know not if the Vicar would pursue this line of research if I myself were not enamored so with the natural world...

I’ve always found this passage delicious for the ominous, dark cloud completely unnoticed by our young journalist. For all of his unusual cleverness, he fails to recognize one of the most disruptive events in the history of industry: the emergence of the textile industry in pre-Victorian Britain. Soon not just his mother but weavers across England, Western Europe, China, India would be forever cast adrift. It almost seems silly to hear people speak today about the automation of weavers creating new jobs, while the artifacts of the era instead note how "the bones of the cotton weavers are bleaching the plains of India." The young man is even more naive because he sees it as an opportunity for his chores to get done for him such that he can focus on collecting more ants.

Annotated Journal. Jahosafats Goosepimpel. c. 1805

...I engaged at the suggestion of the Vicar in a number of lectures carried out by the society for the study of natural history taking place every fortnight in town and after only a dozen such ventures I feel they have broadened my knowledge significantly. I have since left off taking both tea and supper to hasten to the Vicarage in those moments of rest in the day to undertake my studies further among that rich collection of books to which I have no means at home.

Despite encouragement from the Vicar I would never be able to travel into town alone but for Mothers misfortune and that of Mistress Dorothy Moore and M. Pussyfoot. How rude a situation that Mother’s affliction abet my own success and that I learn so much because she has been forced to take service work for which I accompany her near weekly. The Family cannot manage without her additional support, no matter how meager that income most times is which therefore astonishes me that...

It is a wonder Turbish who has easier means to travel than I does not attend such meetings of bright minds when it is of no small inconvenience for me to come from the farm but I do so with such great zeal. I mourn over his faults going uncorrected. I would never have discovered Papilio aegeria if not for my fortuitous attendance at the lectures by Sir Thomas Bootstrap: very common in the lanes leading through woody situations during the whole summer, two or three distinct broods being produced annually...

What a cumbrous reversion it must have been for someone’s sense of self, to this family more generally, to go from a trade to a service. While Master Goosepimpel’s mother and most other women in villages like these were certainly not artisans themselves, they nevertheless contributed significantly to the vitality and economy of small villages across England. The deep class divide between being a servant and having a servant in pre-Victorian England must have felt a chasm to those families falling into the service economy as automation pervaded the nation.

I never managed to find any historical record of Turbish beyond the journals of Master Goosepimpel. I have always imagined him to be a vestigial third or second son of the local gentry, pawned off onto the poor Vicar. Clearly such a young man would have been raised with letters and would have had profoundly better prospects than the brothers Goosepimpel. And yet, it’s fascinating to think of the differences in a family like Turbish’s and a similar family today, the former in their era investing all of their wealth in land while the latter family today commits nearly half its income into the human capital development of their child. Prior to 1830, it didn’t occur to either the wealthy or the poor (so nearly everyone) to invest in human capital; those that could invested in land, not in the professional skill set of their children14.

Jahosafats’ almost fanatical interest in the natural world clearly has nothing to do with investment on the part of his family nor with automation in either agriculture or textiles. It does seem likely, however, that he would never have had free time to engage in these academic pursuits if it were not for the free time automation allowed him, even as his brothers and Turbish failed to seize this freedom.

Annotated Journal. Jahosafats Goosepimpel. c. 1806

...I came across two days hence walking from town a large smatter of the Libellula cancellata: mouth armed with jaws, more than two in number: antennae very thin, filiform, shorter than the thorax: tail furnished with a forked process. The description with which Linneas affords his Libellula cancellata is remarkable for its brevity and as he refers to no other authority the Vicar thinks difficulty might arise to the identity of the insect intended. I however think that the description though concise is expertive and perfectly applicable to the specimen in question.

I’ve had ample time to make account of the changes of these fauna across the seasons due to my daily procession to town on the Vicar’s behalf for which I am quite pleased and by the way of which I further my studies firsthand. I even had the good fortune to show John Malcolm who was greatly delighted in his way. He is a composed little man, his garden is the piece of ground after you go in at the door to the south of the Hut just at the entryway to town...

Jahosafats rambles on for another several paragraphs about the insects in this John Malcolm’s garden before finally returning again, if unintentionally, to the impact of automation on his mother’s labor.

...Mother is not the only weaver left without work. Labor drops even in the city ‘round the bend, local craftsmen increasingly left with naught, no matter how skilled and how beautiful their handywork. There has been a complete hollowing out of all artisan work now. They lose the autonomy of their work and are forced into service to make bread. The town dies as many sit idle and on our own farm there is less demand too for there is less money for goods. Looms go unused o’er the whole country.

As Master Jahosafats precisely describes, the loss of work did not make weavers more creative; they feel febbly downward. They didn’t metamorphosis into greater creativity simply by being “freed” from the industry, and nor did this brilliant and naive boy for being freed from the farm. He was already creative.

Of course, Britain and the West alone saw any benefit at all from such industrialization and jealousy hoarded the benefits. Yet even there, many lost their professions entirely and were plunged into poverty. As I have already noted, countries in the East–India and China most clearly–saw many whither and starve to their deaths.

We have yet to even broach the subject of the child labor that became rampant with the rise of the factory. Countless children of this period fit into the low-level jobs that industrialization created. Artisan weavers were squeezed out of work by small-fingered children whose labor, by nature of their age and vulnerability, was a fraction of the cost. The textile industry hit its peak in the 1850s, but jobs were already being lost to machines as early as the turn of the century. Why did this fracturing society wait for 50 years before investing in literacy specifically and human capital generally?

Annotated Journal. Jahosafats Goosepimpel. c. 1807

I have received recommendation in London for clerical work from the Vicar as well as introduction by letter to Mr. James Pumpernickel, author of some writings on the study of insects to which I’ve taken to with great ardour. I intend to tell that prolific scholar of my findings on Philaena lineataria: antennae gradually tapering the base to the tip: tongue spiral: wings in general deflected when at rest: pale anterior wings with an oblique bilineated band at the base.

This specimen I am inclined to confide is a very rare species indeed and will be observed to differ from that species previously concluded by my teacher as the Philaena lineataria. I should be sorry to sow the seeds of discord with one that has done such a service to me but none of my findings agree with my Master’s on these moths in particular but I nevertheless anticipate Mr. Pumpernickel will be quite taken with...

The cruel threshold for the spontaneous creative is dropping precipitously as a function of automation. Over the long generations leading up to the Industrial Revolution, the number of creatives slowly increased with the population, but there was not yet interest in cultural transmission of creativity through educational attainment. The rich owned things; they didn’t make things. So, spontaneous creatives quickly began to dominate innovative fields, forcing a degree of economic inclusivity we still see today15. Of course, this inclusivity and increased productivity predominantly benefited the wealthy regardless, because they owned everything.

The men at the lectures in the city suggest that with the power loom and mills comes great progress for the economy of England. Try telling that to Mother. And yet I admit that I do not doubt in the progress of the economy, for I myself wish to grow in it, through my studies and the work that I have cultivated for the Vicar. The town’s dead, and I’m leaving. London provides opportunities that are nonexistent here. Those men of lectures are excited about such opportunities, as am I of course, and yet they do not see what’s happened to those left behind from this religion of progress.

I see not why my brothers and mother and those left without work could not be considered advantageous to that progress too. We do not waste coal or yarn so why waste my brothers?

I desperately desire for more and suffered year after year of lashing for my seeming cheek but such lashing paid off in more than full as it gives me the chance to further my study and to prove myself. I would be left to working on the farm unable to feed myself if I had done as my brothers, or forced to find work elsewhere off at the coal mines dying of explosions and consumption. Why did they not follow suit and move forward with me when the chance was readily upon them? I try to lift myself and see a vivid necessity for change–why must they try and hinder my passions so?

Jahosafats’s greatest insight wasn’t about insects, it was about the almost complete absence of investments in human capital at this time–not by families, not by factories, not by countries. And yet, the young journalist, though intelligent, is still oblivious: women couldn’t have done what he did. Historically, men may move laterally, but women always seem to move down to make space for others, and only a courageous few move up into the creative economy spontaneously.

Why is progress about machines and not about people? When I say progress, I mean developments in an economic sense. Johasavat is right in the middle of this period. Unfortunately for him his revelations were thirty years too early; if he actually lived to retirement, it wouldn’t be until roughly 1830 that the rest of Britain caught up with his perspective on the value of literacy.


“That’s what’s so great about AI: it always makes people more creative.”
– Pompous Mansplainer

Freedom ≠ Creative

The creative economy is defined by our ability to explore the unknown. Routine skills like writing boilerplate code or reviewing contracts for loopholes, however cognitively complex, simply don't make people more valuable in the creative economy. Beyond even what we do with the skills we have, it is our ability to learn and adapt to new problems and environments that makes us creative.

The creative economy is home to many of the most lauded aspects of human existence: inventions, art, science, innovation. And yet, for its outsized position in history, the creative economy actually plays a fairly insignificant role in the lives of most people. For most of us, our economic life is a job, one in which we are told what to do and paid in accordance with some concrete set of skills. Whether you are folding t-shirts at a clothing store or writing code for your company’s app, the vast majority of us are part of the service economy. All routine labor–cognitive or non-cognitive, low-skill or high-skill–is part of this service economy. Some economists, slaves to formalisms that they are, refer to this as non-non-routine cognitive labor. This is stupid. Nonetheless, let’s refer to workers in the service economy as non-nons16.

No job is wholly segmented into creative or service. Jobs are complex combinations of individual tasks ranging from the lowest-skilled routine labor to rarefied explorations of the unknown. Yet, even the most professional jobs are still predominantly made up of routine tasks.

If some jackass somewhere with a fancy title tells you what to do, congratulations, you are a non-non in the service economy. On the other side, if people look at you and shake their heads and say, “Poor bastard is married to her work,” you might be in the creative economy17. Could someone with a similar background, education, and skills step into your job immediately, or is your value to the economy18 you?

Within the traditional paradigm of cognitive and non-cognitive labor, education is the solution. It is the engine that drives emerging economies towards the promised land of high-value knowledge and sustains innovative ecosystems. But if we consider the economy fundamentally split between creative and service, can education truly make people more creative?

Writing Wrongs

I founded my first company, Augniscient, to end high-stakes testing. Finals? Dead. SAT? Dead. MCAT, GRE, LSAT? All dead. If a student is spending weeks taking adderall and cramming for it, it should be axed. Instead, the act of learning itself should be the assessment. Augniscient’s modest proposal was to end high-stakes testing and replace it with actual learning.

Through their everyday work, students produce a huge volume of what education researchers call “artifacts”: conversations, homeworks, and the other everyday byproducts of the classroom environment. The vast majority of this is entirely ephemeral, contributing nothing specific to teachers’ or parents’ understanding of the progress of students or the eventual assessment of their knowledge. I invented a machine learning approach that could accumulate clues from these artifacts and derive insights into the conceptual understanding of each student. As I envisioned it, the best part was that you never needed to give them an exam. By “listening” to their freeform conversations in online discussion forums, my co-authors and I found that we could actually predict students’ final grades starting in the first week of a course. As the class continued, our models were able to predict with greater and greater accuracy what they would get wrong or right on their final exams. So why have a final exam at all?

Our main finding wasn’t that we could predict students’ grades19, but rather what predicted students’ grades. Unsurprisingly, our system predicted poor outcomes for students that spent large amounts of time either off topic or discussing genuinely incorrect ideas. More interestingly, our system strongly preferred evidence-based reasoning over anecdotal thinking. For example, in an undergraduate biology course, students would frequently say things like “I remember this time when my dog…” during a discussion about evolution. Needless to say, no students’ observations about their dog’s behavior are going to be particularly useful in thinking about the debates and theories of a process typically measured over millions of years. The AI quickly learned to flag such phrases as negative predictors of course outcomes.

Right or wrong, transforming what they had learned and expanding the framework of the curriculum was a better predictor of success than simply following the rules.

I was thrilled that our “cognitive analytics” algorithm could identify the types of phrases one might assume are related to student outcomes, but its most interesting finding was a shock: being right in the moment was often not a positive predictor of learning. It turned out that students that accurately and normatively reproduced ideas from the course were frequently predicted to perform worse by our model. In contrast, students that explored ideas beyond the materials, even if they were occasionally wrong, were predicted to perform better. For example, many MBA students in an Introduction to Economics class would accurately discuss the basics of collective bargaining in nursing. They weren’t wrong, and yet our algorithm predicted slightly worse grades for those individuals. Traditional assessments give full credit to right answers, yet our system learned that these students were hiding their own misconcepts by quoting the rote material they’d learned in the class. Other students that took the ideas of collective bargaining in directions beyond the course material, such as by auto workers, were predicted to perform better, even if they were occasionally factually wrong. Right or wrong, transforming what they had learned and expanding the framework of the curriculum was a better predictor of success than simply following the rules.

If your entire career depends on a list of skills and knowledges on which you scored highly in school, then why should we be surprised that students never want to be wrong in class? An economist might say that all of these incentives favor performance over growth, but in the case of these discussion forums, the instructors’ assessment of a student’s participation was simply a check-box—if they posted twice in a week they received full credit. There was no formal incentive to be right; everyone earned credit for participation either way. Despite this relative freedom, the majority of students feared appearing wrong more than they were motivated to explore new ideas. But it was those few students willing to risk a “dumb question” in the short term that expanded their understanding of the material in the long term.

This is exactly what the creative economy is all about: exploring the unknown. Creatives translate their knowledge into new spaces and push boundaries. Clearly an education system focused solely on knowledge-retention and rote performance does not make students more creative. But those students risking dumb questions also learned that trying a new idea and being wrong, rather than simply reflecting their shortcomings, supported growth and learning20. Our education system must set aside its obsession with routine skills, however complex, and instead embrace a formal curriculum of exploration.

Of course, that’s a lot harder than it seems, and I should know.

Stingebot

At my second education company, Socos Learning, we wanted to explore how decisions made in the classroom affected long-term life outcomes after graduation. To do this, we started a bit more modestly by building a machine learning tool called QuickStep to help university educators link individual student experiences in the class not to a grade, but to that student’s job placement a year after graduation. QuickStep offered a dashboard showing how educator-student interactions–sentiment of feedback, frequency of interaction, responsiveness to intervention–affected students’ outcomes. With this dashboard, I envisioned teachers personalizing their support for each individual student, turning them, in a sense, into data scientist-teachers.

...we found that the right word at the right time to the right student had an outsized effect.

QuickStep tracked variables that had nothing to do with traditional education measures, skills, or knowledges. We looked at everything else about students’ lives (ranging from raw numbers like days needed to complete an assignment or quiz scores to more abstract transformations like sentiment or self-assessment) and found that some of these variables were hugely predictive of success in class but rarely generalized to everyone21. Still, all of these messy clues began to give targeted insights into who a student could be and how to get them there. Importantly, we found that the right word at the right time to the right student had an outsized effect.

As QuickStep snooped for clues amongst all the myriad variables associated with student identity22, experiences, and performance, it revealed some truly astounding connections between non-academic experiences and long-term outcomes, but that data was frequently still so limited. For example, in one university’s continuing education program, entering students were prompted to write an essay on what they would do with a $1 million award. Conceptually, the essay was meant to reveal the nuanced strengths and weaknesses of each student entering the program. In reality, the students wrote almost nothing. Nearly every student stopped after two sentences, with the first almost always a restatement of the question. What was meant to be a revealing artifact was instead a complete waste of time for the students and the educators.

As a general rule, I hate chatbots. They are the lowest of the low of AI nonsense. However, even chatbots have their moments, and for Socos Learning this was it. We built a specialized chatbot, Stingebot, that would draw out the students during the process of writing their essays. It began with the very simple question: “In one sentence, what would you do with $1 million that’s different from everyone else in the world?”

It turns out students’ responses were still depressingly stereotyped, but at least they had a touch more flavor. They’d spend money on health issues or taxes, cars or homes, and that little bit of nuance gave Stingebot some leverage. The system applied a semantic analysis to the students’ responses, identified the concept that was most distinct from other students, and followed up with a further prompt that was simultaneously a subtle attention reward: “How interesting. You’d spend it on cancer research. Why is that?”23

Students would respond to that additional prompt, and Stingebot would in turn identify distinguishing features in their new responses to give further prompts. The process was clunky and sometimes awkward, but in the end students effectively wrote essays of 3-4 paragraphs. Not only did that give the QuickStep system more to work with, but Stingebot itself was designed with an internal model of student reward learning. Many of these older learners didn’t have long careers of academic success or reason to feel that their hard work on this essay would pay off. Dopaminergic circuits in their brains were likely wildly signaling missed rewards every moment they were working on this damn essay instead of doing literally anything else in their lives.

Rather than expecting the students to produce an entire essay from scratch after many years away from school, Stingebot provided its own proximal reward in the form of attention. Instead of an all-or-nothing distal reward for writing an entire essay, our bot provided breadcrumb rewards that lead a student towards a more complete essay. It even learned the students’ reward schedule, delaying its replies longer and longer in an effort to scaffold each student’s motivation to continue. (Learning and leveraging reward schedules is a bit like The Price is Right: get as close to their actual reward delay without going over. Wait too long and the signal turns from reward to unlearning.)

Stingebot used an extremely modest algorithm; it was neither a sophisticated deep neural network like Microsoft’s Tay nor was it a racist, sexist wreck of an AI like Microsoft’s Tay. Still, turning the process of writing an essay into a conversation resulted unambiguously in longer and richer essays. Even with its limitations, Stingebot demonstrated how education could directly scaffold creativity in students. Where cognitive analytics revealed the potential of students that were willing to explore, Stingebot and QuickStep showed that any student’s creativity can be directly engaged.

...what makes a person creative is an expression of what makes them different from everyone else, and the courage and opportunity to share it with the world.

By design, our bot’s scaffolding didn’t introduce any new ideas into students’ essays. It simply helped uncover more of who the students already were. This suggests an interesting possibility: what makes you creative is you. In other words, what makes a person creative is an expression of what makes them different from everyone else, and the courage and opportunity to share it with the world.

While our work with students suggests that with effort a curriculum for creativity is possible, many thought leaders seem to think that automation has always made people more creative, and AI will inevitably do the same. For the sake of QuickStep, Stingebot, and nearly half of my previous companies, I wish those jackasses weren’t so completely and utterly wrong.

I Invent Broccoli

I’ve been proud of nearly every product that my companies have released, and my work in education most of all. We published scientific papers, gave invited talks, and presented demos around the country with the belief that we would transform teaching. But every teacher that played with Augniscient’s cognitive analytics system or saw a demo of QuickStep’s dashboard said the same thing: “That’s cool…and a little terrifying. And what the hell am I supposed to do with it?” We imagined that we were handing teachers a tool to influence the life outcomes of their students. With AI augmenting these teachers, I believed (and still do) that we could transform teaching into a fundamentally more creative and intimate job24.

Wonderful though that aspiration might sound, it does not fit with teaching as it exists today. We built a system perfect for some parallel universe in which all educators are Master Teachers with nothing but free time on their hands. When actual teachers saw our demos, most had no idea what to do with the bloated mass of information that our systems shoved in their faces. After all, most educators already work 70 hours a week but only get paid for 4025. The whole point was to augment teachers, but just like most other professions, teachers were profoundly resistant to changing their processes. They found our dashboard arcane, overwhelming, and menacing.

There has been a long history of revolutionary educational technology entering the market with astonishing promise...and virtually no effect. At Carnegie Learning, a team of incredibly intelligent and motivated scientists spent years in the lab developing cognitive tutors. These sophisticated AI systems could infer what kids were thinking when they made mistakes in order to give targeted hints that supported conceptual growth. In numerous trials, the cognitive tutors accelerated students’ algebra learning, but when similar EdTech products actually launched into schools, most had modest to no effect or actually decreased student performance.

Many teachers in classrooms using these technologies feel like they are being demoted from teacher to teaching assistant, deposed by the sales pitch for a shiny new computer. Those of us developing these systems imagined we were empowering teachers to take control of their classrooms. Knewton was one of the first companies to bring artificial intelligence into a student-facing curriculum with their “mind-reading robo tutor in the sky”. On visits to their New York offices I was impressed by the machine learning team...and then came Knewton’s notorious pilot at Arizona State University. Initially, the leaders of both Knewton and ASU praised the “successful” pilot but consistently refused to release the actual details. When it was finally revealed that most classrooms performed worse using Knewton’s tutors, its CEO26 blamed the teachers27.

This disconnect between technological aspirations and classroom realities also plays out in the schools themselves. “Technology first” charter schools have proliferated, placing students in front of computers for much of the day. Many of these classroom-based technologies make teachers feel like they’ve been automated away, and research indicates that the vast majority of technology in the classroom actually reduces students' outcomes. This obsession with a utopian zeitgeist of technology solving all problems has failed to make students more creative or even improve their grades.

Each of these innovative products had huge potential to help and could have genuinely been a foundation for a more creative learning experience for both teachers and students. But in every case, they (and I) simply assumed that the presence of the technology would inevitably lead to better outcomes. For all its AI sophistication, cognitive analytics never made teachers or students more creative, even when they were given the freedom to explore without negative consequences. QuickStep offered unique insights into long-term outcomes for students, but most teachers saw that as a distraction from their day-to-day job. The AI tutors from Carnegie Learning and Knewton, along with MOOCs and flipped classrooms, were revolutionary innovations meant to transform a generation of students. All of these technologies, including my own, were responding to the same basic impulse: because we can imagine a world in which these technologies do good, that world is inevitable.

Sadly, it doesn’t work that way. The idea that any technology will make people more creative simply by existing is ludicrous. The vast majority of people, educators included, are heavily entrenched in a pattern of routine labor and systems that discourage creativity. Shoving technology into their hands and saying “go” will not transform work from non-non into creative overnight.

There were amazing teachers in this space that were making use of our technology in inspiring ways. It was incredible to see what they did with it, and the hope was that our technology would help all teachers be just as creative. I imagined QuickStep would give every teacher the power to nudge the long-term outcomes of their students. Though QuickStep never quite achieved this, economist Raj Chetty’s research has identified that a small number of such teachers already exist. Ironically, their students tend to slightly underperform on standardized assessments but still have a significantly higher likelihood of attending university and landing better jobs. These teachers are imparting something deeper to their students than routine academic performance, something considerably more creative. Their students are like those revealed by our cognitive analytics systems, exploring the unknown despite the incentives.

Related research at the level of schools has shown a similar disconnect between testing of rote skills and improvement of long-term outcomes, possibly mediated by peer role modeling. The students that pushed the boundaries of the curriculum had the best outcomes. In fact, the only one of these technologies that seemed to play an active role in increasingly student creativity was the meager Stingebot. And the one thing that differentiated Stingebot from the rest was our explicit intention for it to push students’ boundaries. The only technology that actually made a difference (rock on, Stingebot!) is the only thing that was designed intentionally for the purpose of promoting creativity.

Robo-Recruiter28

...if you actually want to make positive change, innovation must focus not on what’s imaginable but what’s inevitable in the real world.

I understand all this now but at the time I felt confused and frustrated by the response to QuickStep. When the teachers “didn’t know what to do with it”, I was baffled. “Do you have no vision? No sense of imagination?” But that wasn’t the point. I simply couldn’t expect teachers, or anyone, to take an entirely new tool that didn’t have a clear relationship to their current work and do amazing things with it. Most approaches to innovation are simply imagining what’s possible in an ideal world, but if you actually want to make positive change, innovation must focus not on what’s imaginable but what’s inevitable in the real world. I do believe that teachers are capable of doing amazing things with these technologies but I also understand why the vast majority won’t. They’re just like everyone else.

Recruiters are also just like everyone else. Despite the AI-driven tools we built at Gild (and the large sum that companies paid to access them), nearly every job search using Gild’s product still started with recruiters running keyword searches for specific job skills against our massive databases of candidates. If programming is fundamentally a routine skill29, then recruiting or assessing people purely based on skills makes sense. You either know how to do the job or you don’t. Our research at Gild showed that this wasn’t true. Even for jobs mired in routine labor, the factors associated with creative labor differentiated not only quality of work but also long-term life-outcomes. Even when in non-non roles, creative individuals do amazing things.

For 100 years recruiters have been looking at name, school, and last job as a secret code that reveals the best candidates. As the body of research revealing the flaws in this thinking grew, companies demanded alternatives and happily paid Gild thousands of dollars per license to apply our fancy tools. And what did their recruiters do with our AI? They ran a skill search for their city, and then looked at name, school, and last job, just as they had done for the past hundred years.

Still, there were a handful of individuals using our product in amazing and creative ways. Gild had a tool to auto template personalized letters, and the tiny number of recruiters that put it to use had amazing success. (In fact, our own internal recruiting team used it to find developers in Detroit30 to teach an unpaid six-week course for the nonprofit, Girls Who Code. Nearly 75% of developers that received our letters responded, and a quarter took the unpaid “job”.)

The talent executives spent money on our vision of what recruiting could be, and we had actual evidence of how systems like this could make a difference. Despite all of the potential, giving AI tools to recruiters did not change how the vast majority did their jobs. It most certainly did not make them more creative. Thought leaders keep selling us the idea that AI will automate away all the rote, boring work and free us all to become creatives. But when these recruiters were given the chance to realize that vision, nearly all of them used our product in the exact same way they used every other tool.

I’ve seen how exceptional teachers, students, and recruiters transformed their work using the tools described above. It’s easy to understand the blind hope that just making technology available would drive everyone to do this kind of work. Chetty’s research showed that the best teachers taught deeper life skills that didn’t appear on any test. We know from our own work with cognitive analytics that the best students are those that explore, make mistakes, and try new things. The research we did at Gild showed that the best recruiters are willing to go the extra mile to understand what makes a great employee. All of the tools we built were designed to help individuals make a creative leap in their work, but putting those tools in their hands is only one piece of the puzzle. We saw that when given the tools and a choice, most people overwhelmingly chose to continue their non-non work.

My research with Stingebot found intriguing evidence that evoking and developing creativity really is possible, but experiences at Gild and across numerous EdTech projects demonstrated a brutal truth: the idea that technology will magically empower remains pure myth. Of course people can change31, but that change comes from intentional effort. It is not the inevitable result of some Econ101 supply and demand curve.

In The Second Machine Age32, Brynjolfsson33 and McAfee describe a world in which “...people with connected smartphones or tablets anywhere in the world have access to many (if not most) of the same communication resources and information that we do while sitting in our offices at MIT. They can search the Web and browse Wikipedia. They can follow online courses, some of them taught by the best in the academic world. They can share their insights on blogs, Facebook, Twitter, and many other services, most of which are free. They can even conduct sophisticated data analyses using cloud resources such as Amazon Web Services and R, an open source application for statistics. In short, they can be full contributors in the work of innovation and knowledge creation.”

Bull. Shit. Do you really think any kid, phone in hand, would spend their Sunday running R scripts on AWS? Alright, my kids might be weird enough to do this, and I might have done it too if it had existed, but doesn’t that point out how privileged such a choice would be. The possibility of that scenario does not translate to its inevitability.

And yet, what if all kids did choose to “be full contributors”? What would it take to get them there? We know what doesn’t:

  • What doesn’t produce creatives is our current education system, university or not.
  • What doesn’t hire for creativity are our current hiring practices.
  • What doesn’t train for exploration are six week upskilling programs.

So where the hell do creatives come from?

The collected works of Prof. Schlashenschlessen

“I’m not worried–the wheel, the plow–new innovation always means more jobs for horses.”
– Glue

“Instead came inventions that made the horse obsolete—the tractor, the car, and the tank. After tractors rolled onto American farms in the early 20th century, the population of horses and mules began to decline steeply, falling nearly 50 percent by the 1930s and 90 percent by the 1950s.”
– Derek Thompson,
The Atlantic

From The Collected Letters of Archibald Moore, August 6, 1890

Dearly Beloved Parents,

I write with the most excellent news that Grover and I have arrived safely in the Steel City by river boat. Our journey over was filled with excitement. Grover expressed some great elation in talks of making the city ours. He spoke with fervor of the money to be made in the industry of this great city and envisioned us to be the next Carnegies! Such talks reminded me of his animated daydreaming back in Ohio, keeping me in good company daily while I measured and manipulated the machinery of the farm after supper.

Pittsburgh is glorious! The city’s expanse seems endless. She is dotted all over with marvelous feats of architecture in bridges and engineering in streetcars and carriages. Great factories line the river with fields of smokestacks belching constant steam and smoke that sting the eye but smell of progress.

Because we are strong and able we have some small jobs running crates while in search of work. I am quite enamored with the thought that I shall bring great change in America. I already see much to be improved.

Accomodation proved tricky and we slept on a local inn floor for some nights before finding room for rent for some 70 cents a week. The room is small and cold and hard – the view of the cityscape though a hellish sight still astounds. We watch as armies of factory laborers ride the funicular daily up and down the hillside like ants. The Incline from above makes my heart race. Another wonder of engineering!

Dear Father, it is my sincerest hope that life on the farm is improved to some degree with the work that I did on the Oldsmar engine before my departure. I hope now to apply myself with the new opportunities that this great city should afford. It is my hope too to save a sum if only to send to you and Mother.

Yours,
Archie

I simply adore the youthful eagerness of Grover depicted in these early letters from Archie. Imagine them, bright and young, seemingly on the cusp of grand adventure, unaware of the wildly different paths their lives will take. I tingle with anticipation34! Those early nights of Grover passionately gesticulating as Archie tinkers sets the mode for the rest of their lives.

Just as in Archie’s descriptions of Pittsburgh, we see similar industry taking root throughout the Northeast of America in parallel with the great cities of northwest Europe. Outside these grand metropolises, much of the world is still reeling from the economic collapse wrought by the Industrial Revolution, with only Japan emerging as an industrial power. All of these regions share one drive in common with Pittsburgh: a seeming insatiable hunger for routine factory labor.

Reading through these letters, I see how this hunger consumed Grover even as it lured him on with promises of a glorious life, a dismal story of lateral transition35 for those like him. The transition from the farm to the steel factory was not necessarily better, though that steel did make railroads and in effect transform our country.

From The Collected Letters of Archibald Moore, December 17, 1890

Dearly Beloved Parents,

I write with the most excellent news that Grover and I have arrived safely in the Steel City by river boat.

I received your letter today in good health, which I wish for you too, from God, from the depths of my heart. I thrill to hear that the flood did not reach the farm and rejoice at news of Margie’s baby and wish her excellent health.

The weather in Pittsburgh turns cold and in a town with no single patch of flat land the icy streets become a spectacle. The trees lay bare and there is occasional snow though grey and brown by midday. The fires of the steel mills are stark, the icy ground reflecting every so often the burning red from the foundries.

I have had some good fortune where work is concerned. I learned shortly after fixing the wagon of a man on the street it seems unbeknownst to me at the time I had aided somewhat serendipitously a man of some importance! I have since been offered an apprenticeship with the engineers of one of the largest factories about town. It is rewarding work as my fingers itch to explore further and I am delighted with the opportunity to improve these great machines, to return to the examinations and explorations I pursued home in Ohio, though now of course I am surrounded by machines of wondrous proportion and capability increased tenfold. Indeed, the work I have accomplished already to that line has brought some small notice from the manager and an additional sum of pay which I have attached to this letter, Dear Parents.

Though I am delighted with my position I can’t help but notice the conditions of the factory for those who work on the line. Men from a proliferation of countries come to Pittsburgh to work in steel. The work of the city is brutal and hard and often stretches some sixteen hours daily. Here they select workers just as they pick out beasts at the market – just as long as they are strong and healthy; that is how they deal with people in the city. But it is true, that if one is strong, young, healthy, and industrious, then pay is regular.

I have seen Grover’s ambition to work and make money. He is young, healthy, industrious. He waits in line eagerly and works long and hard and he gets his work much like on the farm. He makes his eight dollars a week. But Grover spends twelve hours daily in exhausting and consuming labor like an automata devoid of impulse. Is it surprising that his remaining hours are spent in either sleep or play? I wish that I could show him my work tending to these great machines, as I once did back in the barn at home, so he could see how his labor contributes to the great transformation of America, but he tires so by the last bell and really must rest. I do understand that for Grover the notion of giving another four hours daily to the brutalist Frick is a horror.

Here we see the formerly towering threshold for spontaneous creativity lowered further: indefatigable Archie finds the opportunity to create his own circumstances in a moment of roadside charity while seeking his fortune in the city. Unlike in the case of Goosepimpel, who was lashed brutally for his absence from the fields, the advent of agricultural automation raised productivity such that Archie’s presence on the farm was no longer vital. He was thus freed to follow the flows of his own endogenous motivation, tinkering, building, bettering.

I imagine that Archie represents all of the fantasies of Industrial Revolution apologists: a world in which, freed from the farms, every worker becomes an artist or captain of industry. But in fact, we still see all the sacrifices Archie describes on his path to success—the endless unpaid hours tinkering. The dominant story is still clearly that of poor Grover, a head flush with lustrous dreams but days filled with soot-stained labor even more routine than a day on the farm.

I can only imagine that his weariness of work would have been mine, had I rushed to the line alongside Grover. He tried to pull me along with him saying such things like, “This is the best way to become involved in steel. To become the next Carnegie!” But then good fortune tipped the carriage that day, bringing newfound opportunity. I now spend my twelve hours daily most engaged and put in four more with sometimes never having even noticed the clock!

Certainly at times I do envy the freedom of those on the line for the feminine company they keep – the factory, for all its diversions, is quiet in the night, with nothing but the sounds of my own tinkering for company. Grover seems near marriage with a half dozen women these days. I spend my day with the workers and engineers of the factory and the evenings with naught but this pesky boiler.

I would still not trade my nightly battles with these violent boilers for I am confident that my labors will come to bear greater fruits yet and this excites me greatly, Dear Parents. I know my friend pities me some, that I am always under fire for my work, but I cannot stop myself– I must save this factory! For it has burned not once but twice these past six months and four men have died. I cannot also forget that we had dreams of owning and running the factory together.

I luxuriate in these glorious letters as I see a boy immersed tirelessly in his work. We see in Archibald’s passion the spark of another spontaneous creative. He stays late to work without being asked and makes sacrifices without obvious return to himself. Toying and tinkering with technology since he was a child on the farm, little did the boy know that he would find himself involved in changing the face of the modern factory!

The experience of young Archibald evokes Crompton and Carnegie, as well as our own Master Jahosafats Goosepimpel. The early lives of all of these creatives were spent immersed in the very worlds in which they would make their greatest contributions. Carnegie began running telegraph messages in Pittsburgh and engrossed himself in the dealings of business leaders37. Returning back even further to Britain, we have the example of Samuel Crompton, who began spinning with a Jenny as a boy, and was among those–not dissimilar to young Archie–who set out to improve it, leading to the famed Crompton Mule of 177936.

For some months I have been frustrated with the limited power of the boilers of the steam engine in the factory basement. I know that if we could only change its pressure capacity that we could boost our progress. My work at night has afforded me the time to partake in my own private experiments and try these models and the results are quite exceptional I must say. Alas they don’t listen for the assumption that a farm boy knows nothing of engineering.

It seems those from the higher families in Pittsburgh with their posh accent are not inclined to take a farmhand from Ohio with equal seriousness. But oh, how narrow they are! None of them have read of the latest discoveries from Darmstardt and Berlin. If only they had interest in pushing that boundary further! I would have them understand that my modest circumstances would not hold me back. I remain determined and I begin to see those such restraints against my aspirations already waning.

By this era, we see new mechanisms for the transmission of status, not merely via wealth, but also through the transmission of capacity. While it’s clear that intrepid Archie has increased existential breathing room to explore, it is at the same time hard for this earthy Ohio farm lad to enter a creative economy at the cusp of being dominated by cultural transmission. The available space in this emerging economy slowly fills with those inheriting human capital capacity from their parents. One can almost feel new spurious signals of culture and creativity–university and social network–joining with traditional signals, like the right family name, that overshadow the actual qualities that make a creative.

I trust well that you have received the sum I sent home last month. I pray that it may be put to good use toward the new John Deere Father hopes to buy in good time to more easily and efficiently tend the farm.

Yours,
Archie

From The Collected Letters of Archibald Moore, November 21, 1892

Dear Mummsie,

I received your last letter and have delayed so in writing to you for I tire greatly by day’s end but I pray you are safe and well.

News from Pittsburgh is near same as always. I’ve traded waking up at dawn and feeding the animals and pulling the plough to waking up at dawn and pulling a lever and moving crates. I do not miss the uncertainty of the seasons in the fields but I do miss the sunshine on my face – often days go by when I do not see the sun for I wake before sunrise and return home so long after dark. The shifts run together day after day without a seam near repetitive as can be, dull, rhythmic, repetitive, the factory floor even following me into sleep.

I do not dare complain of the money I have been able to earn that I might send home to you, Mother, but I cannot help but lament the fact that for so many years I have done the same work on the line. If I had known that I would have been doing the same work every day to put money in someone else’s pocket, I daresay I would have just stayed sharecropping with you and Father. I have tried at this work for so many years. It seems there is no moving forward. The factory hires new immigrants that speak an English barely our own that work just as well but complain not for wages nor conditions for they otherwise starve in their Country.

Yes, dear Mother, I had heard of Archie’s tour of New York touring with his damned boiler. There has been no explosion since his invention. Many among the factory workers are grateful for his work though some feel forgotten by one of ours. I think often of our journey up the river with aspirations for a great life. Over these years we both have worked diligently, though of different work, Archie stumbling on great fortune that put him in company with great men. I may never have had a knack for boilers but given a bit of Archie’s luck I would be touring the world too.

Best regards to my sisters and brothers, friends and relatives, and all acquaintances. Warmest wishes to the Moores and please share my congratulations for Archie’s success.

Your son,
Grover

What a find! This gem written by Mr. Grover Samuels to his mother, a friend and neighbor of the Moore family, was fortuitously captured within the bail of collected letters from Archibald Moore to his family in Ohio. Were they shared with the Moore’s to curry favor? In any case, a remarkable find!

I cannot help, in reading the differing language of Grover and Archie, but reflect on how little sense of agency Grover has compared to his friend. He wallows in his belief that there is nothing that he could have done to change his luck. Comparing his romantic aspirations to his plodding fatalism in Pittsburgh, it’s as if he hopes to win the lottery having never even bought a ticket.

From The Collected Letters of Archibald Moore, May 25, 1893

Dearly Beloved Parents,

I take my pen to tell you the good news. I am safe and sound and arrived on the continent across the Atlantic, thank God, some four weeks ago. Well, it is a good sailing time now. I received your letter, in which I learned that my sister Margie is again with child. What excellent news!

So much has happened since my arrival which I shall update you here upon. I pray that you forgive the delay in my informing you that I had arrived safely to Europe which I hope you will understand when you hear the reason.

News archives place Archie in Hamburg, Germany, in the spring of 1893. What follows in his series of letters from this trip is a compendium of observations on the industrial functions of the great cities of Europe compiled over the course of many weeks. Relevant to my most pressing research are his descriptions of diverse encounters with the inventors he meets in these cities. Who knew that Archie had the heart of a travel writer? What surely must have been meant as a series of letters turned into a rambling account of his revelations as he stormed from one city to the next to demonstrate his boiler.

I had three most beautiful days in Constatt in Stuttgart also where I met a Gottlieb Daimler, a most prolific inventor and responsible for the high-speed gasoline engine we see slowly introduced in America. I understood with my conversation with Daimler that he actually began as the son of a baker… he told me that my mechanical tour of the great European nations paralleled one he did when he was my age!

Daimler was the son of a master baker and intended by his parents to become a municipal employee for the city of Schorndorf, Germany. For me, Daimler exemplifies the emerging class of knowledge workers driven by the aspirations of parents in terms of education and employment. These are the foundations of the38 cultural transmission of the creative class we see today.

It amuses me that his family only succeeded in transmitting the aspirational value of education and capacity-building. Despite their best efforts, young Gottlieb had no interest in becoming a civil servant and pursued an interest in mechanical engineering against his family’s wishes. Much like Archie himself and many others described in Archie’s letters, Gottlieb found and pursued his passions with vigor from a young age. Daimler apprenticed himself to a gunsmith for several years before training as a mechanical engineer at a steam-engine factory and later completing a degree in mechanical engineering at Stuttgart Polytechnic.

I was later made acquainted with a man by the name of Rudolf Diesel whose engine achieves some of the very highest thermal efficiency of any combustion engine under development, and yet he was the child of immigrants from Bavaria that grew up in a leather shop pushing barrows full of animal parts through the streets of Paris and yet went on to receive scholarships and academic awards across Europe.

Poor young Diesel faced repeated challenges in his life: war, poverty, and repeated separation from his family. Still, Rudolf was the son of a professional. That role modeling combined with an obvious touch of genius resulted in a young man that, despite his youth pushing around a wheelbarrow, found himself inclined to build an entirely new kind of engine.

...It feels I near simultaneously met one Karl Benz who has had great success in the construction of the horseless carriage in recent years and reaching near an astonishing three horse power that could leave Cleveland after breakfast and reach Pittsburgh before supper with not one horse. How fortunate we all are to live in such a glorious age! Though I expect it will be some time before we see these Benz motorwagen in America, I remain most inspired by the strides taken by these great men. He even confides in me that he prepares for the first motorwagen competition. I see great prospects on the horizon for us. Such brilliance, these European engineers!

Of Archie’s three new teutonic chums, Benz certainly represents the best example of a spontaneous creative. Benz was raised in poverty by a single mother having lost his father at the tender age of two. His mother advocated for his education, but he was not exposed daily to the professional life of the day. He was, instead, a prodigy like Master Goosepimple. Note how different these men are from each other. It’s obvious from the diverse backgrounds of these men that there is no magic recipe for creativity.

Next on Archie’s itinerary are the glories of France at the height of the Belle Époque! How I envy him…

...no sooner I had arrived in Toulouse when I made the acquaintance of one Clément Ader, an inventor who dreams of putting men in the air! An Édouard Michelin in Paris spoke of great success in filling bicycles tires with air. Once an artist, Michelin began his work in engineering when...These French are an airy folk!

The Tour Eiffel is truly a wonder of engineering, dear mother, I do wish that you could see its beauty and marvel at its height. When the sun sets it is illuminated by hundreds of gas lamps–a sublime sight, from both below and from its highest point. The view of the city is marvelous.

Note again how different these inventors are, wide-ranging interests and histories. It’s clear from Archie’s continued encounters with the most renowned technical minds of the modern world that he was experiencing the heyday of the inventor. Daimler, Michelin, Benz...all became household names. Something about this moment in history was necessary, though still rarely sufficient, to allow for these spontaneous creatives to impact the world. Think of those described in his letters who came from particularly limited circumstances. It is likely that someone like Ader, born into a modest family of woodworkers, would not have excelled to have such an impact on the world of aviation had he been born a century before. Astonishing wealth inequality almost defines the Gilded Age, but the world had changed. The soaring productivity gains of factory line automation and scientific discovery had lowered the threshold for entry into the creative economy. What heights could Master Goosepimple have achieved had he been born just these 90 years later?

Yet let us not fool ourselves. It would take another 90 years for these opportunities to reach much of the rest of the world. The threshold was indeed lowered, but only for a select few in select places.

Upon my arrival on the Isles I met one James Blyth who has just received a patent for something one calls a "wind engine". How extraordinary! He, like me, is the son of a farmer; he won a scholarship to study in Edinburgh and now travels the continent to teach and share his knowledges. I also meet his family, a charming wife… every day he brings his children to the factory to explore the machines. I see how Mr. Blyth has created something and is excited to pass it onto his children. There is such a value in making such an investment in one’s children at such a young age.

Archie’s domestic interactions with the family Blyth reveal this era’s inventors’ growing investment in the human capital of their own children. The coming wars will drain much of the inherited wealth of Europe and allow this next generation, inheriting capacity rather than capitol alone, to dominate the world. Archie appears understandably oblivious to the emerging effects of such strides in cultural transmission, most notably that it would act to obstruct spontaneous creatives such as himself in future generations.

It seems I was mistaken dear mother–for the Frenchman Michelin did not invent the pneumatic tire but improved upon the design of some Scottish inventor John Dunlop that has worked towards incorporating such tires into bicycle racing. Both, still, contribute greatly to this field.

...

The city of London is greater than I had ever dreamed a city could be–a train, powered by electricity, runs under the ground, if you will believe it.

Although I am not yet suffering any homesickness, I do look forward to applying what I have learned from these most excellent men back at work in America and sharing all that I have learned in my travels with the factory in Pittsburgh. The singular genius of the Europeans has stoked the flame of my passion for these works.

Upon my arrival on the Isles I met one James Blyth who has just received a patent for something one calls a "wind engine". How extraordinary! He, like me, is the son of a farmer; he won a scholarship to study in Edinburgh and now travels the continent to teach and share his knowledges. I also meet his family, a charming wife… every day he brings his children to the factory to explore the machines. I see how Mr. Blyth has created something and is excited to pass it onto his children. There is such a value in making such an investment in one’s children at such a young age.

There is a series of patterns emerging among these creatives, but none are universal. Where do these spontaneous creatives come from, if not as a direct result of the automation that lowered the threshold for entry?

Life has been so good to me during these three years since leaving the farm. I have more than I ever imagined. As the accountant of a local newspaper man said to me, "Well, it's no trick to make a lot of money... if what you want to do is make a lot of money." I did not really understand this at first. I have since realized that it has never been my particular dream to make money and I never look to make wealth or spend it; I simply wanted to build a better boiler. I see this in my peers the other inventors here in Europe. There is an obvious difference to me now in aspiring to create wealth and aspiring to be wealthy.

...

I am sitting now at the Savoy in London looking out on the River Thames and I see a fresh and eager young man off a boat coming into the city. Did you know I asked Grover to journey with me to Europe and see the great cities, to meet the men of measure we idolized? He said, “I understand why such a trip might be exciting to you, Archie, but nothing for me in Europe is worth the trip.”

Faithfully yours,
Archie

The parallel here: nothing–not free time, not more money–actually changes people’s lives; only the people themselves can change their fate39. I repeat that industrialization was necessary but not sufficient for a low threshold for spontaneous creatives. Independent of the productivity gains of the Industrial Revolution that preceded them, these men were all on a path to invention of some distinction.

Read Part 3