Doing things poorly

Against the high-tech classroom

On the fictional planet of Arbre, from Neal Stephenson’s Anathem, society has been divided into two highly separate castes. Thousands of years before the book’s setting, a nuclear catastrophe persuaded the population that the world could not survive the marriage of science and technology. Thereafter, the planet’s scientists and intellectuals were confined to highly segregated monastic communities and required, under severe penalties, to never touch advanced technology and to confine their scientific and intellectual pursuits to pure 'theorics.' Regular society, denuded of all but its technicians, froze in time, numbed by mild drugs in the food supply and captured by ubiquitous entertainment.

Life went on on Arbre for over three thousand years, punctuated by controlled outings for the cloistered 'avout' in one, ten, one hundred, or one thousand year intervals, depending on the monastic order these intellectuals belonged to. Without spoiling much more of the book, I will only say that the 'Sæcular Power' suddenly had an urgent need to collaborate with the secluded theoreticians, eroding the strict boundary between theorics and praxis.

When Fraa Erasmas, a cast-off child brought at a young age to the concent, is summoned, along with many of his fellow monks, to consult with the secular powers, he comes for the first time in ten years into contact with the technological gizmos, devices, and machines in the outside world. Stephenson largely borrows from our own world, slapping new names onto familiar technologies: the Internet (Reticulum); the smartphone (jeejah); the movies (speelies). But “Raz” has now spent ten years totally cut off from these devices and their delights. Instead, he has been steeped in an intensely technical, but analog, system of education. He is accustomed to performing long and arduous mathematical calculations by hand; studying and debating at length with his mentors the competing theories of mind and matter that percolate through the concents; and reading hand-copied manuscripts dating back thousands of years describing the finer points of the human genome and the movement of celestial bodies.

His first contact with the outside world is an accident: on Apert, the ten day period every ten years when his 'Math' is able to open their gates and exchange information with and visit the outside world, Raz looks for his half-sister, Cord. He finds her operating an incredibly sophisticated (and incredibly illegal, for him) piece of 'praxic' equipment to manufacture a component needed for his concent's enormous, hand-wound clock. Spellbound, he watches with dawning understanding as the machine manipulates matter whose properties he knows by heart, using geometric transformations he recognizes, to produce a complex shape.

This is only the first glimpse he has of the praxic technology so dangerous for him to combine with his broad theoretical knowledge. What the two siblings, isolated for ten years, have discovered, is two ends of the same string. And it is no great loss to Raz that his education so far has so far been utterly devoid of these praxic shortcuts. He is accustomed to long problems: some of the ones he works on in the concent are hundreds or thousands of years in the solving. But it is only outside the walls, when he feels the pressure of time as his adventure unfolds, that the speed at which these praxic devices can complete tasks becomes important for him. And then these jeejaws and Reticulae are not devices for producing knowledge, but only tools for calculating results or acquiring information.

By radically separating the Mathic World and the Sæculum, Stephenson gives us a practically non-existent scenario as a thought experiment: at its purest, what is the point of reading, writing, and arithmetic? Why do we learn long division when we all have calculators? Why read the book when you can review the Cliff's Notes summary on demand? Why write the essay when you can have ChatGPT write it for you?

When all Raz cared about were the results of a calculation or information about some topic of interest, all he needed was a fully-charged jeejaw with access to the Reticulum. But it would have been impossible to follow Anathem if Stephenson had begun his book 'extramuros.' Visiting the concent, and laboring over its long and (admittedly, sometimes tedious) disquisitions on fictional philosophical and theoretical topics, revealed to us how Raz was able to reason his way through the situations he encountered.

As adults, we have grown so used to the technological shortcuts and scaffolding around us that we have forgotten how we developed a conceptual intuition for the beauty of mathematics, or the magic of the written and spoken word. Knowledge is not facts. Knowledge unfolds within us and transforms us through the slow labor and quiet joy of learning. Education works not by telling us the answer, or teaching us the technical shortcuts for arriving at the answer, but by teaching us how any answer comes about

We risk falling victim to our own success. The brightest minds of a generation are tackling challenging technical problems, like AI. Reading, writing, and arithmetic are all at risk: reading and writing, through the opaque churn of the LLM, and arithmetic, through frankly incredible tools like WolframAlpha.

Who will pick up these challenges in the future? Learning to read, write, and manipulate numbers do not teach technical skills or convey results: they are hard, arduous, thankless, and painful journeys that every student must wrestle with. It is the wrestling, and the hard-won understanding that arises from it, that imparts knowledge. The use value of the output is not important. What matters is the development of the thinker, and the uses to which they will put their hard-won experience in the future.

But without new and bright human minds brought again and again to the toughest intellectual challenges of the day, every effort to cure cancer or visit the stars will wane and dim, as the ladder of inquiry is kicked out from under each subsequent generation. A few months ago, I wrote the following, out of concern that we don’t think about the developmental impacts of social media or AI:

We know some traits that are useful no matter the milieu: problem-solving ability, internal locus of control, grit, creativity. These skills will help make someone more 'successful' as a user of social media—and successful in other areas of life. But how did they acquire those skills in the first place? Social media and the habits that attend it can be corrosive all of these traits by its very nature. Especially because we are discussing the impact of social media on children, we must ask whether social media itself might have the effect of draining the pool of Cowen's Emergent Ventures prospects, by inundating young minds with the kinds of easy and fleeting sensations that are inimical to the development of patience, determination, and delayed gratification.¹

And two months ago, in defense of making things worse than what you could buy in a store:

Tinkering also develops a series of small but meaningful competencies in the tinkerer. This does not mean becoming an expert in all of the fields of science and technology that one might touch. But mediocrity, or even making things worse or more basic than a consumer product is, carries with it no shame. It is not pointless because the outcome is worse; tinkering develops a deep curiosity for how things work that has far longer-lasting impact than how well some homebrew device functions in the near term.²

It is safe to say that it has been on my mind.

Who is doing it well?

The roots of education as we know it in the Western world originate over the course of thousands of years. We sometimes call them the ‘liberal arts.’ From the classical quadrivium (the scientific arts of arithmetic, astronomy, geometry, and music) and trivium (grammar, logic, and rhetoric), to the Great Books movement, to the vestigial requirements for a liberal arts education in American universities, there has for thousands of years at least some kind of commitment to disciplined and painstaking inquiry heavily focused on reading, writing, speaking, and calculating. In this light, it is better to think of current arguments for the use of intellectually-assistive tools like AI in the classroom (which sometimes read like generic AI-generated marketing copy themselves) as proposing a radical departure from the historical lineage of Western education.

In my own small way, I felt like one of the avout in Anathem during my late nights in architecture studio. My undergraduate university architecture program required us to use mostly analog techniques for drawing and building in the first two years. The countless hours I spent perched over a drafting table, taught me a lot of technical skills I no longer use. But it also gave me a tool for design thinking I still return to. There is a reason architects still sketch, even after school, and even with every advanced tool at their disposal: there is nothing like the most primitive of tools for ‘thinking on paper’.

In Anathem, when Fraa Erasmas made the discovery that made him so invaluable to the Sæcular Power, he did not even have a pen to record his insights. Armed with only a sharp instrument and some careful guesses, he and Suur Ala, another member of the concent, pricked holes in a paper to record the movement of a suspicious celestial body. But it was only their deep understanding of orbital mechanics, the properties of optics, and their careful attention that they noticed a pattern of scattered light on the floor and were able to record it so accurately. If I were there, I’m not sure I would have even noticed.

1.

2.