Is There a Traitor in The Country of the Blind?
Seldom is it that I am compelled to write about a book before I can confidently say I am done with it. I want to know what I’m talking about, I need to know what I’m talking about. Half-formed thoughts on art don’t create nuance nor meaningful discourse on a work’s place in society. Even more seldom is it that the progenitor of some of my ire is reasonably likely to at least interact with the piece of writing they inspired. Hi Andrew! I saw you followed me on Instagram and don’t you worry, I’m nice.
Andrew Leland’s The Country of the Blind came to me with a strong recommendation from the same bookseller at Birchbark Books that also recommended Sorrowland. You should be aware of how that turned out.
Andrew has retinitis pigmentosa, just like me, and is a writer, just like me (albeit far more successful), and a teacher, just like me (albeit probably paid more). If the divinity responsible for dropping the cast iron ball of disappointment on my foot with Sorrowland had the capacity for regret, it went about rectifying its mishap with The Country of the Blind. It just didn’t notice the monkey paw twitch in the background.
If there is one point that I attempt to get my students to remember it is that something can not be for everyone. There is always an audience and the better you can define your audience, the better you can understand your audience, the better you can write. On the flip of that coin I also tell them that they will not be the audience for everything they read and sometimes they’ll have to power through that and still try to find common ground anyway. Or at least attempt to cosplay as a member of the intended audience so they can have a healthy understanding of the text.
My initial suspicion that I wasn’t Andrew’s intended audience came early. “My hope is that this book will encourage the sighted reader to likewise discover the largely invisible terrain of blindness, as well as other ways of living and thinking they might not have previously considered.” (Introduction: The End Begins xxv) I am not a sighted reader and that alienated me. Not just the sentence, but many of the objectively interesting ideas Andrew has written about so far fall flat to me because, well, they’re common in my life. I, too, held shame about using my cane for eight years. I, too, sometimes feel like a fraud around both my Blind comrades and the pre-disabled riff-raff that stumble around. I, too, have had to quickly adapt from a visual lifestyle to a more tactile one (Andrew and I actually differ a touch here as my rp is a bit more aggressive than his. Get rekt Andrew, I’m better than you).
There’s a relief in being seen by someone else, but an entire book predicated on the notion of educating someone on disability is potentially wasted on someone who is, well, educated on disability. Our similarities may likely be the main driver behind my current unease about the diegetic text. (Diegetic will become important later, so sit tight). I adore reading texts by authors outside of the Blind wheelhouse as I adore learning about the sheer variety of disabled culture. Andrew talks about exactly this with an entire page dedicated to describing the unique linguistic culture and experiences had by Deaf folks. His website also mentions writing about Protactile, a language I learned about through DeafBlind poet John Lee Clark.
But as I said, I am one chapter and an introduction in and I need a lot more to definitively stick by this prognosis, but then why am I writing about this?
Well, the text gives me some non-diegetic concerns, some kill the author shenanigans that gave me pause. As with all non-fiction books I read I kept Andrew Leland firmly alive through the entire reading experience. I went to his website and other publications and Instagram and, oh, wait, what’s that? Two videos, a pair of clips from an interview he did with Aboard. He’s….praising generative artificial intelligence? Aboard is a company that leverages A.I. assets to improve scalable deliverables for innovative companies on both sides of the Mississippi?
Andrew. What the fuck? “How can I trust you,” I said out loud, “now that I know you’re one of those?”
But I am a dutiful scholar who, after stomping about and raging about this discovery, listened to the entire podcast episode while playing the new map released today on Battlefield 6. Which is the only way I can comfortably listen to three white men talk about large language models without losing my mind.
I’m going to be succinctly honest: the only reason I am not putting this book down for good is because Andrew Leland seems a sound lad who 1) is looking in the wrong places for solutions to problems we all have and 2) got inappropriately clipped out of context.
Andrew is praising Claude Code specifically in the way it allowed a handful of disabled people to create tools that were both unavailable to them and outside of their technical skill level to develop. The extent of Andrew’s “vibe coding” enthusiasm comes from the potential these tools offer to that very specific sect of folks AND he admits his own shame in using said tools as he actively hides his usage of them from his wife who, like me, would see the entire system “burned down.”
The two podcast hosts, self described capitalists, couldn’t help but push that idea far beyond what, I believe, Andrew is comfortable with. The last half of the podcast is Andrew occasionally telling an interesting accessibility story and then getting railroaded into silence by two tech bros who are desperate to show generative A.I. as a messianic figure to humanity. Which doesn’t fully redeem Andrew in my eyes as he actually contradicts a lot of his own points throughout the episode, but that only tempers my rage down to something more akin to enthusiastic scepticism.
Dia duit Andrew, I hope you’re still here. Andrew talks, at length, about an amazing community of Blind hacktivists who enthusiastically produce accessibility tools that are outside of the capitalistic scope of interest most companies have. He pairs this with an anecdote about someone he knew who asked a friend to build a tool that would allow him to scroll on a computer, but the tool didn’t quite live up to expectations and so he used Claude to code a tool that worked. He also talked about someone who used Claude to create a tool that made Wikipedia’s browsing experience more friendly to screen readers as they parse tables of contents and “skimmable” materials.
Do you notice the contradiction here? Claude is presented as a sort of “last mile” solution to these problems, but the true last mile solution was already there and has been the entire time. Claude is, by design, actively concealing the community of Blind programmers and technology enthusiasts who are dedicated to creating these tools. The scrolling tool didn’t work completely as expected. Yea, that happens, that’s called drafting. A rough first draft isn’t made better by running it through a robot, but by asking around and sharing it with friends and collaborating with the folks who know how to build these tools. Why would I vibe code an idea when I can seek my comrades and compel us all into learning and expressing new skills in its development?
Of course the primary issue here is capital. Who wants to do any of this shit for free? Which, yea, fair. Blind hacktivists gotta eat like the rest of us, but Claude is not free. It requires both payment from you and enacts an incredible toll on our communities, ecology, and economy. I am not one to suggest that the solution to our woes is to expect people on the internet to do free labour for us, but open sourcing projects and inspiring a collective working environment splits up a workload into manageable chunks. The money someone would spend on Claude Code can easily go towards a donation to anyone with the technical skills and interest to work on your project. And when a human being drinks a gallon of water I know that gallon is going to be pissed out by someone with hopes and aspirations. Someone capable of love and laughter and other cliché, kitsch ass shit.
When a data centre hosting a portion of Claude Code guzzles a gallon of water it pollutes a community composed of the very people that this technology is claiming to help. It sucks up that water and spits out a poison of the air that drives folks to migraines and illness. It devours our data that’s then sold to the highest bidder and used to further the interests of technocratic overlords who laugh atop a stack of money that could, mathematically, reach the fucking moon.
During Andrew’s prolonged silences the two hosts of the podcast would wax poetic about the “end of the product era,” and talk about how these tools will put solutions into the hands of the individual and ensure that someone can make their own creation personalised to them. Then they slip in the but. How, instead of relying on a corporation to build a tool that works for you, you can get paired with a person, “like a clinician,” and have them observe your life until they understand how to prompt the model to generate the tool you specifically need. The implication being for money. The people who want large language models and generative artificial intelligence to succeed may dress up their motivations in glittering cloaks of accessibility and individualism, but their core is always monetisation. They didn’t listen to Andrew’s anecdotes and think, “Nice! Look at how these tools are used for good. I wonder how we can make this more accessible and ethical in response to the common critiques we get?” They, like all capitalists do, thought, “Look at this market of people who I can convince need this tool. Look at all of these cripples who need to be convinced that their woes aren’t solved by community, but by a rugged, fragile, plastic individualism that I offer them through the monetisation and distribution of my unsustainable product.”
I know this not only because I can smell the motivations of a capitalist from three galaxies over, but because while one of the hosts was passionately talking to Andrew about the virtues of older technology, and how Blind folks are finding new uses for programs that are outdated in the eyes of current paradigm, the other host was constantly begging the subject to be brought to how A.I. is helping disabled people. The first half of the podcast is a marvel in disabled ingenuity and how we have a tendency to craft new tools out of existing materials due to an intense necessity, you know, human ingenuity. But the constant pestering reminder plastered through the transcript is that we’re all on this podcast to sell an A.I. product, not talk about the collective intelligence of the Crippled community.
I am going to continue reading The Country of the Blind. I hope that I finish it and that I write a sequel to this essay where I confidently say that Andrew Leland is essential reading to anyone compelled to learn more about the Blind community. I’ll come on here, my humble platform, and say, “You know, I don’t agree with Andrew using Claude Code, and the instant I catch a whiff of him using a large language model to write something I’ll tell all of my students that he is a traitor to his comrades, but damn did he write a fine book.”
Anyway, hi Andrew, hope you read the whole thing. Go ahead and share this to everyone you know and all of your friends and your generally larger-than-mine sphere of influence and say something funny like, “Ope. I would have gotten away with it if it weren’t for you meddling cripples!”