Stretching an analogy real far
Mathematicians are like Luddites. Yes, the actual weavers, not just the slur against people who are against technology. They were weavers in general: weavers, knitters, and cloth finishers defending the value of their skill and lifestyle.
Etymology
Maybe astrology for men, but just as fun.
The Proto-Indo European root tek-/tekō/*tek/: to plait/to weave.
Text: words and letters woven together.
Context: that which is woven together (con-*).
Textile. Subtle: the underlying (sub) thread.
Subtext. not etymologically a duplicate of “subtle” but sure feels like one.
Technology: from the word tekne, which means “craft”. Weaving is an old craft, about 20,000 years old. Flint knapping and fire making are older, and damned hard, but weaving has a combination of freedom and skill.
We talk about ideas as though we were handling fabric because, for a very long time, making fabric was one of our most sophisticated skills .
History of Weavers
Master weavers in the Tudor era had to work crazy fast. I don’t recall the exact number, but it’s a combination of speed and skill that isn’t really matched today. Especially in the speed aspect, done by hand.
The raw material, (the text) of weaving, has always been really really time consuming to make. And producing it was necessary to do anything further with it.
Until the Jacquard loom and its cousins, ironically maybe the first true programs. So the computer came for weavers before computers even existed, and Jacquard inspired Babbage.
The Luddites were skilled weavers. The automated looms suddenly collapsed one aspect of their skill, previously inextricably tied to the production of the raw material.
They did not react well to that disappearing, and sought to destroy the machines. But their position in society now ran against economics and therefore the wheel of progress, and their interests became less and less relevant.
When they started destroying, the response was brutal. Troops came in and shot many of them, as did factory owners.
Later, the whole industry sublated. Because the literal raw material of weaving, the cloth, became so cheap, made to measure and prefabricated (off the rack) clothing became possible and soon common. The design aspect of the art was still valuable. This is basically the province of a fashion designer now. But fashion design scales (that’s what a fashion line is), and most of the jobs disappeared. The few that remained are maybe even more prestigious and better paying, but much fewer.
History of Mathematicians
For most of history, mathematics was a hobby of the rich. Later, when basic arithmetic became more widespread, it gained a lower skilled class: “human computers” and scribes (long) before that. Ordinary computation was relegated to them. Abacuses and the like help diffuse this. But abstract math has always had a big cliff blocking further progress and learning into it.
Mathematicians have always had to do computation, but it was not their self conception of their main job after a point. Especially in the 1900s, the research center of math unmoored itself from most other sciences. Leaving a more and more insular group, essentially high priests and sometimes hierophants.
Until the machines have really taken off recently.
Mechanization in both fields
Leibniz worked on a mechanical calculator in the 1670s. He wanted to automate work so “excellent minds” wouldn’t really need to argue, but just “calculate”. Later, de Prony’s large table-making project divided labor into a few people choosing methods, a supervisor group turning the methods into instructions, and god knows how many performing arithmetic. Babbage described this organization in 1832, explicitly comparing it to a cotton mill.
Mechanization has changed both fields, in similar ways. The raw material of math is the proof, and the raw material of weaving is the cloth. Now both are cheap to make. Definitions are harder to launder, but I expect AI will succeed at that too within a year.
While machines are not generally intelligent, the few that are at the frontier of both industries (esp math because ai is a lot better at the virtual than the physical world right now) have greater position than ever. Certainly they can do way more than before by using these tools.
In any case, both math and weaving have changed.
Weaving underwent its own reorganizations. Patterned cloth required skill long before Jacquard, and there was still a high skill ceiling even as tools improved. His mechanism, demonstrated in 1801 and developed commercially around 1804–05, used punched cards to determine which warp threads rose. btw this is where the gpu warp term comes from. A design could be encoded and repeatedly executed aka mechanized.
The British Luddites did not simply rise against the Jacquard loom. Different textile workers opposed different machines and employment practices. In 1808, before the named Luddite disturbances, a Royton handbill protested Parliament’s rejection of a minimum-wage bill. In Arnold in 1811, framework knitters destroyed 63 frames belonging to unpopular hosiers. Yorkshire croppers were skilled cloth finishers threatened by shearing machinery; Lancashire weavers faced power looms. These people already used machines. The dispute was about which machines, under whose control, on what terms, and with what effect on the value of their work.
Mathematicians have tried to stem the tide in some ways, saying that they didn’t want the machines to do their work. Recently they had some declaration (I’m not gonna look it up) about “human mathematics” but economics is against this desire too. And so progress’s wheel rolls (and crushes). Math, against common mathematician belief, is valuable. Really, really valuable. It was just too damn hard to use usefully (until now). The AI galaxy brain is redeeming math’s commercial value, which is why normal people would care. Normal people interact with math through its fruits, and if you took a math paper and turned it upside down and had them read it, it wouldn’t be any less confusing.
The analogy to fashion designers is that the machine can synthesize or spin up new “cloth” so quickly now, leaving people at least for a while at a higher level.
This loops back to the idea of technology and tekne, the Greek conception of arts, or skills that can be improved. Craft is mechanistic and teleological. The machines can grind through the mechanistic aspect, but its teleology is still up for grabs. For now.
The analogies collapse as AI advances because of its meta-aspect, inserting itself into the workflow at every level. I think the order of what AI masters is the human tech tree, turned backwards. So I expect weaving and fashion design to fall after math, but I very much believe they will. AI will supply its own teleology.
Some people at the very highest end still have a position because they can simply use the tools, especially for math, in a way that others can’t because they can just read the outputs. And hardly anyone can do that. And while the AI might produce stuff that outstrips them, they’re still a lot less outstripped than other people, giving an even smaller group of this already small group an exalted position. But I expect it to pop like a bubble in the end, and for the end to be a very finite amount of time away.
The craft itself is greater than ever, for both. Or more like for 1.5. One of my old obsessions was tailoring, and watching the old tailors die is pretty depressing. Ready-made clothing transformed the mass market, though bespoke tailoring survives, but only for the best.
So, yeah you should be mad as a mathematician. You have (had) a meaningful position with respect to status and salary, at least among people you cared about. The scarcity behind that is dissolving. And that sucks.
More literal and metaphorical cloth will be produced than ever, but it’s rly different.
It’s not likely to ratchet backwards, leaving math in this new state.