What I Read in June 2026
Books
Fiction
Dumb Witness – Agatha Christie
A classic Poirot novel, with the novelty being that he's been hired by a woman who has already been murdered. It's always fun to have Hastings around, plus a dog is heavily featured. Very enjoyable, aside from the attitudes that are rather of their time.
Death on the Nile – Agatha Christie
Another Poirot novel. This one was a very complicated plot, but resolved nicely. As usual with Christie – especially the holiday ones – there were parts that have not aged well, but overall this was very enjoyable.
The Great When – Alan Moore
I picked this up as a "Buy One Get One Half Price" deal in a Scottish Waterstones over a year ago, and just about managed to get round to reading it. I enjoyed it, although it did take me a while to get into.
At first I found Moore's style cloying, but it was not as exhausting to read once past the introduction. After this point I found it rather pleasant, and I found myself reminded of Terry Pratchett.
The narrative was well-structured and satisfying. Despite being filled with fantastical elements and whimsy, the book retains an essentially bleak viewpoint throughout, but the well-wrought resolutions to the many threads keeps the dreariness from feeling overwhelming.
I'm sure that one more well-read than I could have had a lot of fun catching the dizzying quantity of references. Even with my morsel of cultural knowledge I found plenty. I was surprised to learn just while writing this that many characters in the novel were in fact real historical figures, so I probably missed a whole extra dimension of experience. The same is probably true of London geography – my own visits to the capital have been brief and filled with work, so my knowledge of its layout is practically nil.
Appointment with Death – Agatha Christie
I managed to squeeze another Poirot in before the end of the month. This is another holiday novel, which certainly felt apt in the ludicrous heat wave I read it in. This was a more straightforward plot than Death on the Nile, and also shorter. It was a pleasant, standard Poirot. This does, however, also include the standard uncomfortable reading when those of non-European extraction are mentioned.
Non-Fiction
The Mechanic and the Luddite – Jathan Sadowski
This was an excellent read. The book is a collection of essays criticising technology, capitalism, and the relationship between the two. There's also a framing device of two personae that should be taken while performing such criticism: the eponymous Mechanic and Luddite. I found the defence of Luddism delightful, and found the movement strongly aligns to my current views on technology.
I'm trying to write a full review of this book covering all the contents, but my time has been absorbed by preparation for a Pathfinder game I'll be running. In the meantime, my post about Ted Sandyman was based on a rare original thought inspired by the book.
I'd heavily recommend this book to anybody who is interested in technology, and also grows tired of it serving power instead of people.
Extrastatecraft – Keller Easterling
Following on from The Mechanic and the Luddite, I thought I would read some of the references used. This was the first I tried. I found it very dense and abstract, to the point where I could not grasp what it was trying to say. I think this is more an issue of me not being the intended audience, perhaps, than any issue with the book itself. I get the impression it's more of a dense academic text, and it's not in a field I'm familiar with.
The full title is Extrastatecraft: The Power of Infrastructure Space. Alas, by the end I was no nearer understanding either extrastatecraft or infrastructure space. So abstract were the explanations of these that I found the explanation of a "topology" later on much more concrete. I hoped that the context from reading would help me understand, but I was too lost by the time any discussion came in.
The histories of zones, broadband in Kenya, and ISO 9000 are all interesting in their own right, but I cannot fathom what the book was trying to say about them.
Articles
The Material of Male Power – Cynthia Cockburn
I found this in the references of The Mechanic and the Luddite. It was a good read, approachable even for one so unversed in feminist theory as me. Cockburn uses the type composition industry as a case study of the interplay of patriarchy and capitalism, highlighting how male compositors faced with technology transformative to their industry used the little power available to them to keep women out.
One point I thought was particularly strong was about how the male dominance of a workplace ends up being reflected in the new technologies used, and how this can perpetuate the gender dynamic. Type composition is used as an example. Initially composition was a task involving considerable manual labour – type would be placed into a heavy forme which would then be wrestled around the printing shop. As technology progressed, this became less and less necessary, with increasingly arbitrary reasons to exclude women being given. The size and weight of the forme, however, was always arbitrary, right from the beginning. There always could have been women in the composition industry, but the masculine default excluded them.
Another reflection I had, reading about the tactics used to exclude women, was the double edge to the slogan "There is Power in a Union". In this case, that power was often used to maintain high wages for compositors, but also to dominate women.
The paper ends with some speculation as to how the male-dominated industry would react to the latest technological changes of computerised typesetting. This development meant that being a compositor was similar to being a typist, a profession that was dominated by women. One suggestion was that typing would cease to be a gendered activity, which seems to me to be what actually happened. I'm reminded of the gender dynamics of software, which was once seen as "woman's work," but has developed into being imbalanced in the other direction as more men entered the field.
As God Might Be – Meghan O'Gieblyn
Another reference of The Mechanic and the Luddite, this one is a book review of three books on technology, each coming from a different perspective. As such it offers something of an overview on philosophies of technology. It was a pleasant read.
First is the view that technology is inevitable, which seems to be the dominant view in Silicon Valley, judging by the hot air wafting around "AI" discourse. Second is a view with an understanding that people decide what technology is invested in and comes out on top, but lacking any systemic critique of how those choices are made. Finally is a pessimistic view that progress is not a linear process, and seems to focus more on the idea of artificial intelligence necessarily being imbued with human flaws.
I don't think I really agree with any of these. The first I reject outright, it seems to me completely fatalistic to believe this, and I largely hear it touted as an excuse to do whatever the speaker wants. The second is closest to my position: people do decide what technology will be progressed upon – as can be seen by the $1.4T spent on "AI" – but we need to consider the systems behind the decisions instead of asserting that all is down to individuals. The third has my sympathies, in that progress is certainly a non-linear process. I don't agree with any hand-waving around artificial intelligence, however, as one may as well be explaining fundamental truths about unicorns.
Does Improved Technology Mean Progress? – Leo Marx
This was a delightful article I found once more in the bibliography of The Mechanic and the Luddite. Marx runs through a history of the connection between technology and progress, from the enlightenment ideals founding the United States of America to the modern day. It feels somewhat US-centric, but that was Marx's nationality so I won't hold it against the piece.
In the enlightenment science and technology were seen as a means to an end. Improved technology meant progress only so long as it was serving some social goal (such as a revolution against the British). As time went on, however, it became more of an end in itself. Marx cites writings about railroads which remind me of current talk of "AI".
There was a backlash, with figures like Henry David Thoreau given as examples. Thoreau in particular is quoted as saying people are turned into tools for their tools, which reminded me of my recent musings on Ted Sandyman. I don't think it's entirely the same – Thoreau is thinking more holistically, about how by obtaining products we end up committing ourselves to endless upkeep and toil to maintain or further our lifestyles – but I think it's a neat parallel.
Marx seems to agree that technology needs a social goal to count as progress. I must concur. Technology is not an end in itself, and needs to be built for a progressive goal in order to be progress. New methods of authoritarian control are not progressive by mere virtue of being new technology.
Deep Learning is Hitting a Wall – Gary Marcus
Yes, another reference from The Mechanic and the Luddite. In this piece Marcus runs briefly through the history of two competing methodologies in Artificial Intelligence research: deep learning and symbol manipulation. Marcus argues that deep learning alone is hitting a wall, and that combining the two methods may yield more results.
This article was published in March 2022, only a few months before the release of ChatGPT turned the wave of deep learning hype into a tsunami. I found it quite interesting to read a perspective on AI research from the before-times. With the benefit of hindsight, I don't think we can really say that Marcus was incorrect about deep learning's wall-bound course. Sure, GPT4 was a scaling-based improvement on GPT3, but since then we've seen that pre-training is no longer yielding significantly better models.
Marcus is further validated in his call for using the two techniques together by recent history. Since hitting the wall on pre-training the focus has shifted to "harnesses" and "agents". From what I hear of the leaked Claude Code source, it uses more traditional AI techniques in tandem with the LLMs it calls out to. In that light this article seems somewhat prescient.
I suppose this is also a good place to talk about other media I've enjoyed this month. I went to see Albion live, and have listened to little else since. My current earworm is their cover of The Pogues' Bottle of Smoke.
In terms of visual media, I rewatched The Two Towers. There's also a new series of Game Changer on Dropout which has been as fun as ever.