Where are the Luddites in the Culture series?

There are two ways in which science fiction mis-projects technology. The first is to take a single piece or field of tech (AI, interstellar travel, etc.) and push it into the future while inadvertently freezing all others at current levels. This is different from but overlaps with “the big lie”, a writing technique where authors allow themselves one large violation of physics with a promise to leave it intact elsewhere. The second way, which is the subject here, is to assume nearly perfect diffusion of technology. Science fiction societies are homogeneous: it’s hard to find any groups within them (at least those that have been contacted) that opt out of certain parts of the tech tree and go about their lives on a lower subtree; unless I’m mis-remembering, there are no Luddites in the Culture series.

This is odd because in our own society we have good examples of such groups. If we array them along a single axis of “tech level” and put the average smartphone user near the top, beneath this we have digital minimalists, Hasidic Jews, off-gridders, Mennonites, Old Order Amish, and at the bottom the semi-contacted groups like the Sentinelese. The latter perhaps shouldn’t be counted because they don’t have full knowledge of the tech tree (thanks to laws upheld by the Indian government), and don’t have the freedom to choose which parts to adopt.

Science fiction assumes all subtrees will eventually collapse into one, or it simply isn’t imaginative enough to explore how they won’t. But I think the subtrees are durable, that they’ll proliferate instead of merging, and soon the majority of people by choice won’t live on the one that includes the uppermost branches. I’ll take these in reverse order.

We’re already seeing a majority opposition to AI in Western societies, with recent polling summarized neatly by Nilay Patel in his piece on “software brain”. There’s a widening gap between those who don’t use AI at all and those orchestrating symphonies of agents all hours of the night. This is adjacent to a growing gap in health tech, where the Bryan Johnsons of the world are becoming equipped with techne that’s just now starting to verge into transhumanism proper (though broad diffusion of GLP-1s might undercut this gap argument). There actually is a rare parallel here from science fiction in the “Ultras” from Alastair Reynolds’ Revelation Space universe: these are the merchant marine crews of lighthugger ships who are the augmented, heaviest users of advanced technology, much of which isn’t adopted by the general population.

The emerging transhuman-ish subtree in our own universe will keep pushing up the tech axis, built around extreme AI use, the quantified self, and increasingly gene editing/epigenetic reprogramming, embryo selection, and brain implants. But it won’t be popular. Of course there have always been early adopters of new technologies. The difference here is that many advances won’t be integrated by a majority going forward. John Gruber captures this discrepancy well:

You can’t argue that the public always turns against groundbreaking technology. The last two epoch-defining shifts in technology were the smartphone in the 2000s, and the Internet/web in the 1990s. Neither of those moments generated this sort of mainstream popular backlash. I’d say in both of those cases, regular people were optimistically curious. The single most distinctive thing about “AI” today is the vociferous public opposition to it and deeply pessimistic expectations about what it’s going to do.

On the other hand, as the y-axis expands it will stretch the distribution, and the cost of staying one or two steps back drops: more stable equilibria become available lower down. Like the Amish, these new groups will carefully consider which pieces of technology to allow and which to exclude; from the outside it will seem arbitrary and in many cases hypocritical, but that won’t matter to those on the inside. These communities will start to create physical spaces based on their subtrees, some of which will be visited by those above or below as a novelty (we already do this with ren faires etc., except the ren faire workers go home at night). 2026 may be the year of analog, but soon “analog” will schism into dozens of different communities. This will happen along disagreements of where to draw the line, but also by constructing positive identities around a particular tech level. This proliferation won’t necessarily be tied to religious beliefs like many of the current groups are, but some of them will evolve into quasi-religions. I don’t agree with Gary Vaynerchuk that it will be a barbell situation, all-or-nothing; the middle is going to be filled in.

Will the subtrees be durable? We can look to the Amish where the data are quite good: 2025 studies suggests a TFR around 6.11 (varying by conservatism), compared to ~1.6 for non-Amish in the U.S. (2024 CDC/NCHS). The conventional wisdom is that this can’t possibly hold, that we shouldn’t plot straight lines on a graph to project future population sizes. But why not? The 1.6 isn’t going to climb, and why should we think the 6.1 will fall so sharply the Amish eventually shrink and diffuse back into society? I don’t mean to say all new subcultures below the cutting edge will necessarily have higher fertility (there may not be a causal link), just that the Amish are a fairly inconvenient truth if you try to argue a community can’t cut itself off from most of the latest technology and not just survive but flourish.

The other argument for durability is the march of technology itself, alluded to above. If you didn’t let antibiotics or the printing press or spears into your community in the past, you weren’t gonna make it. But today the baseline almost everyone accepts is very high, and people can continue to selectively borrow from higher up where they need to. As I wrote in my review of Against the Machine, there’s an irony in that it’s only possible to safely get off the grid and return to the woods because of how advanced is the unexamined tech that has faded into the background. Members from most of the existing subtrees mentioned above seek out life-saving treatment from higher up when they need to: again, this “hypocrisy” doesn’t matter if your life is at stake.

So more subtrees, many of which will be durable, and the thickest ones no longer include the uppermost branches (by choice this time). Where are the weaknesses in the argument? Collapsing everything down to a single “technology level” is a simplification. If we put Bryan Johnson to work on a manual-labor farm to provide for himself, probably he wouldn’t do as well as the average Amish community member. There’s other axes here that could be explored. It also may be too early to conclude the AI backlash will sustain itself and ossify into connected communities or subcultures. But by the same…token, there simply might not have been time for these to get organized.