Recent reading

This was a whirlwind of a week, I’m about to be on holiday for a while, and I am definitely running low on energy. So instead of trying to think new thoughts right now, here is an annotated list of some recent readings.

Books

• Empire of pain: The secret history of the Sackler dynasty, by Patrick Radden Keefe. A devastating study of incentives (read: money) and how they can blind people to the wider reality they are creating.
• The unaccountability machine: Why big systems make terrible decisions, by Dan Davies. I don’t think I had ever heard the phrase “accountability sink” before reading this, but once you grasp the idea it is hard to avoid seeing it everywhere. Davies uses the concept to link cybernetics to economic theory to monetary policy to management consulting to the current state of the world and why it is so broken. It is a pop-culture explainer, but it does that job very well. Mandy Brown has a very good review.
• The this, by Adam Roberts. A very ambitious and structurally complicated novel about communal intelligence, time travel, evolution, and god. I don’t read much science fiction anymore but this was a nice way to dip back into the genre.
• Careless people: A cautionary tale of power, greed, and lost idealism, by Sarah Wynn-Williams. I only picked this up because Meta tried to block its publication. It is fine, but the picture it paints of the people who dictate so much of how our modern world works is incredibly depressing.
• The beginning of infinity: Explanations that transform the world, by David Deutsch. I picked this up after re-reading Will Myddelton’s essay “Research heresies” (which is excellent). This was an exhilarating read, though I’m not entirely sure I know what to make of it. Deutsch manages to weave together a history of science and thought that seems airtight but is also kind of unbelievable. For all its flaws, the book is still absolutely worth spending time with. David Albert’s review in the New York Times ends with a pretty good description of how reading the book feels:

  Deutsch – notwithstanding his open and anti-authoritarian and altogether admirable ideology of inquiry – is positively bubbling over with inviolable principles: that everything is explicable, that materialist interpretations of history are morally wrong, that “the only uniquely significant thing about humans . . . is our ability to create new explanations,” and on and on. And if the reader turns to Pages 64 and 65, she will find illustrations depicting two of them, literally, carved in stone. I swear.

  Never mind. He is exactly who he is, and he is well worth getting to know, and we are very lucky indeed to have him.

  Deutsch’s refrain about “good explanations” felt like it hit on something central to the practice of synthesising user research.
• Ingrid Caven: A novel, by Jean-Jacques Schuhl. I mentioned this fever dream of a “novel” (so far as I can tell it is probably mostly just about true?) a few weeks ago. The prose style was hard going at first, but once I settled into the book’s rhythm the text produces a hallucinatory feeling that is similar to Jon Fosse’s Septology. Highly recommended for fans of Fassbinder, Yves Saint Laurent, and 1970s glamour.

Articles

• When IA fails. Dan Brown on the large variety of problems information architects get asked to solve. This is a useful reference document for introducing people to the work of an IA. That said, I’ve not encountered the kind of resistance to this work that he describes early on – I find IA to be a really tangible thing for most people!
• Bad people do the thing you love. Adrian Howard with a nice little reminder to not demonise other professions because you aren’t getting your way.
• The economics of shared digital infrastructure. A policy paper from UCL’s Institute for Innovation and Public Purpose on treating digital platforms as a form of public infrastructure. There isn’t anything in here that seems controversial to me, but I’m probably not the intended audience. The most important part is the implications for approaches to structuring public sector funding.
• OKRs for Design Orgs. Defining and measuring design “quality” is really hard and I am forever looking for tips on how to do this better. Here, Peter Merholz provides some guidance.
• Monthnote: March 2025. James Higgott (the NHS App’s head of product), with a round up of what’s happened in March at NHS England and the NHS App, which was a very quiet month lol.
• Introducing the “push notification” state (Via James). This gets at something I mentioned in a weeknote not too long ago about the gov maxim of “do the hard work to make things simple”, notably: does it just need to look simple or does it actually need to be simple? A lot of what public sector design and development work does is mask complexity, but over time that complexity compounds and things get unwieldy. You build yourself a technological house of cards.
• ‘Sacrificial lamb’: the fall of NHS England. Fascinating and kind of brutal insider accounting of the past few months in NHS org news. The bit that stands out to me is that I keep hearing “This isn’t about replacing NHSE micromanagement with departmental micromanagement. We are taking power in order to give it away, with resources and responsibilities devolved down to the front line. That’s how to get more innovation, efficiency and better services for patients”, but I have yet to see evidence that this is actually true. How and why does this work? No one ever explains it, and never mind how devolution is often directly at odds with digital transformation.
• Why design goes wrong and how to set it right. Pavel Samsonov has a newsletter, which is nice because it means I can look at Linkedin less often. This is the first post in a four part series about the state of design today. It features a huge number of links to articles from across the web and is worth reading for that reason alone, but Samsonov also manages to weave a nice narrative about the challenges facing the industry and what we all might do to surmount them.
• Grady Booch on Design. Jake Bloom works through Grady Booch’s answer to “what is design?”, and extends it with what Bloom calls a “temporo-material perspective”. This is serious theory nerd stuff, and a kind of writing I don’t encounter much of these days, but it is a lovely rumination on the nature of design work.
• Browser choice is an accessibility consideration. Beeps on why, no, you shouldn’t block Chromium users.
• Comic Sans and the art of imperfection. To quote Sebastian quoting Phil Baines, “there is no such thing as a bad typeface; just bad uses”. The fact that Comic Sans turns out to be rather accessible for people with dyslexia or other reading difficulties always makes me chuckle.
• Markdown and the slow fade of the formatting fetish. This has a nice history of word processing, but I don’t really buy the core argument – in my normal text editor (Obsidian) I use Markdown very specifically to add visual formatting to text, not avoid it. Adding headings, or italicising words, or highlighting sentences is something I do so that the text makes sense at a glance. I’m not doing this for some future publishing project. Further, while I don’t really mind Markdown most of the time (and I got very used to it when I was a heavy user of iA Writer), I really don’t understand why in 2025 I need to look at all this stupid formatting gunk. Shouldn’t computers have figured out how to have a nice WYSIWYG editor without producing totally janky files in the process?
• UX designers must own design judgment. Jonathan Korman with a useful reference point and polemic about the differences between shared responsibility and trusting subject matter experts (be they designers or developers or product managers or whoever).
• How to improve public technology markets taking a systems approach in local planning. Matt Wood-Hill with lessons learned while working to shape the local authority planning space. There is a lot here that rhymes with the challenges I see in the health sector. I think we (the NHS) could learn from the ecosystem approach described.