Newest thoughts →

May 2026

Had a great weekend chatting with @ernestopriego.com about scholarly publishing, intellectual lineages, Pier Paolo Pasolini, the merits and drawback of multiple authorship, worldwide cultures of said multi-authorship, a potential comics data project and more...

Now, I'm off to do some writing of my own and some programming. Journal article to finish this month, then I'm on to writing the Dark Web book.

Having guests does exhaust me, but it also stimulates me and makes me want to write and read and think.

I will sleep well, tonight.

A photo of Ernesto Priego, Martin Eve, Helen Eve, and Lucy Morris at the seaside in Broadstairs

Some people assume that, because I am Very Unwell™ right now, I am incapable of work. It is true that I decided not to teach anymore. But my research is flourishing and I put in full hours & more running Knowledge Commons. I do programming work for DQ Programming, too. But some people try to...

decide for me. They see me and think I don't have the capacity. Or they decide, on my behalf, that I need rest and so can't do something. But I don't need this protection.

So I now invoke: nothing about us, without us.

If you are looking at someone with a disability and deciding, in your head, whether they are capable of something - particularly work - and then excluding them, you are being ableist.

I will tell you, upfront, if I can't do something. And I'm quite a good judge of that now.

This is the main news as far as I'm concerned. PLOS moving to the @janewayolh.bsky.social platform that I founded, with @ajrbyers.bsky.social, for submissions and contributing work upstream to the main project!

Where PLOS is headed This is not a vision PLOS holds at arm’s length. We are building toward it through concrete commitments. PLOS is taking a deliberate approach to developing and contributing to open, shared infrastructure as part of its role in the open science ecosystem. This means not only using open source software, but also actively contributing to it, developing code and shared components, and working with others to shape common solutions. Our transition to Janeway for submissions is one example of this approach. Rather than acting as a passive user of a platform, we are investing engineering effort upstream, contributing code and ideas, and helping to shape the direction of the project alongside a wider community. This approach requires a different way of working. It means engaging with shared standards, working in the open, and accepting the …

I am eating a massive lactose-free Galaxy chocolate bar as a celebration of book news. I am going to keep re-using this celebration discourse all through the weekend.

Also, still my 40th birthday "season".

Went to cite Science Europe's Action Plan for Diamond Open Access. Zotero picked up Science, Europe as the author.

The names Science.
Europe Science.

*CUE BOND THEME*

YES!

It is SUCH a relief when good peer review news comes in! It looks like this book is almost there! Just a few more revisions!

I've received peer review feedback on a book. Second round. Late birthday present?

I won't read until the morning as I find peer review feedback to be viscerally painful. The reader - same as last round - was encyclopedically knowledgeable but I tried to address all their comments.

So we will see.

Round 1
Reader 1: positive
Reader 2: critical (justified)

Round 2
Reader 3: positive
Reader 2: ??? (in my inbox)

I brace myself for tomorrow.

As nights went on and nothing happened [...], most had difficulty remembering the earlier rise of heart, the sense of overture and possibility and went back once again to seeking only orgasm, hallucination, stupor, sleep, to fetch them through the night and prepare them against the day.

In the UK, there is a company structure called a limited liability partnership (LLP) that is easy to set up, runs on a partnership basis, and makes tax management easier.

However, the problem is that they are treated as second-class citizens. There's no digital accounting submission for example.

Most frustrating.

A talented friend sends me a birthday card etching that she did herself. Blown away by the talent and the generosity. "Oh, I'll just knock up an amazing etching for your birthday card, not even framed".

A card with an etching of a forest and trees by Caroline Matthews

Fluke -- my all-time favourite electronic act -- have a new 2x vinyl LP coming out that are reworks and remasters of their classics. There are some audio samples here and I am excited! https://www.forcedexposure.com/Catalog/fluke-the-second-bite-2lp/SR.2601LP.html

The whole Fluke group are electronic musical geniuses as far as I am concerned. Oto is such a classic and very sophisticated, but they seem not to like the later Puppy, which, as far as I am concerned, is an absolute tour-de-force. Switch/Twitch is just unparalleled and Snapshot is amazing.

Can't believe they don't rate that album so highly, even though they were on major Hollywood soundtracks like The Matrix Reloaded.

Anyway, the tracks are out of order here; wrong tracks on wrong titles. but I have enjoyed the samples so much!

I am here at the doctors' surgery, now waiting anxiously to find out how the skin graft is looking and to see whether it's "taking".

The skin graft looks good!

Secret confession: I use voice transcription all the time rather than typing for messages, but I dislike it when people leave actual voice messages. This is kind of selfish because obviously it's more personal and should be nice to hear from friends in this way. But I find hearing difficult.

I seem to have the curse of the big birthdays. Just before my 30th, I had a stroke.

Today, on my 40th, I'm not well and will do haemodialysis. But the sun's out, I'm off work, and I'll celebrate as much as possible. 🎉☺️

I read and greatly enjoyed Shakespeare Was a Woman and Other Heresies: How Doubting the Bard Became the Biggest Taboo in Literature by Elizabeth Winkler. Obviously, her title is bombastic and designed to provoke - but it's not her core argument. I *was* though expecting a blowback.

I wanted to search literary journals for a damning review or systematic debunking etc. It looks like everyone's gone for a lofty/dignified silence. I saw this and haven't read it all yet, but it doesn't look like a savaging https://muse.jhu.edu/pub/140/article/990007

I guess that Stratfordian critics regard this as a trashy, popular book that doesn't properly engage with the scholarship and isn't considered part of the record. it seems a shame to me. This could be an excellent opportunity to systematically work through the authorship questions.

But then I'm being naive. There have been whole books written about this on all sides, and it's hardly going to be resolved by a trade book prompting a debate and/or debunking.

As for me, I am an agnostic. I'm happy to think that the question is not adequately answered by any current theory.

I do object to the 'anybody who questions this is a lunatic' narrative, though. There are some lunatics who question it, but there are also many lunatics who believe the Stratford theory. It is a shame that the alternative authorship discourse has been filled with such conspiracy theories.

But there are things that are not lunatic. I think the fundamental problem boils down to a complete absence of evidence. Evidence of absence is not the same as absence of evidence, I know.

Hmmdeehmm. Bank holiday musings.

I am wondering if anyone in English Lit would like to co-author a paper? That sounds vague, but the project is a digital assessment of writing over the course of our discipline in the 20th century. So looking at what lit academics published in journals and how their writing has changed over time.

Specifically, I am looking for someone or someones who is/are willing to score the jargonicity of 250 short passages. This will give the machine model baseline examples scored with inter-rater reliability.

Then, however much involvement you'd like in the actual writing up is up to you.

I can't at this point give a 100% guarantee of the viability of the classifier that I'm building. Obviously, the whole project depends upon that working reliably, but if you're willing to come on board and take that risk with me, it would be superb.

Please drop me a line, either on here or via email to martin@eve.gd.

The classifiers I am building combine:

• Jargon lexicon terms per 1,000 words
• Rare/coined terms per 1,000 words
• Hyphenated compounds per 1,000 words
• Nominalisations per 1,000 words
...

• Acronyms/initialisms per 1,000 words
• Mean sentence length
• Ratio of long sentences
• Mean noun-chunk length
• Flesch–Kincaid grade level

Each passage gets an absolute score from 0–100.

Articles are then aggregated.

The system keeps all evidence spans, including matched jargon terms, rare/coined words, hyphenated formations, acronyms, and complex sentences, so every score can be audited. This can be published.

I don't know whether this is bonkers or a great project. But I need to do the scoring - with someone else - to know whether it will work. Also, I'm using deterministic classifiers so that it's reproducible.

What it might reveal will probably not be of great shock to many academics. I can predict the trends that we might see using this classifier. But it would be interesting to have such a classifier that we know works and that could be transferred to other disciplines and made generalisable.

Plus, it's a really geeky, fun meta-project about our discipline. It is also an opportunity to make a digital study into something normalised in the discipline by writing it in a way that is accessible to the discipline itself.

So rather than write this up as a scientific paper about the classifier and the mathematical methods and so on, I would like to be able to tell a story about our discipline's evolution that uses these data as evidence.

Your question raises the fundamental challenge of the idea. It's not implementing the technology to do these classifications (although that has some substantial challenges of its own); it's defining a definition of jargon with which a disciplinary audience might agree.

Of course, it could just be that they fall over at the sight of a Freud citation (or whatever stereotype we want to project).

Revisiting NICE's rejection discussion of Evusheld:

"the committee considered that, as shielding reduces, the risk of infection and severe outcomes will increase"

So: we can't give vulnerable people this drug, b/c they'll stop shielding & have a life, then might get Covid.

I find this maddening.

Then they set an impossible standard:

"None of the clinical studies included evidence of efficacy against variants around at the time of this evaluation because of the rapidly evolving nature of the SARS‑CoV‑2 virus."

They took a year to review it! Then complained that studies weren't current.

In vitro studies showed that Evusheld was not efficacious by the time NICE got to it. Their process was too slow.

But I object to that medicalising control over people with disabilities. Deciding whether another group can have treatment that lets them go outside b/c they might get sick if they do.

I acknowledge also that they were modelling direct utility gain (stopping shielding) based on a perceived efficacy of 10%, which was where it was at. So I see there might have been real risk.

A better question: why didn't AstraZeneca model this? They seemed unprepared before the committee.

The more I read about the CDL case, and the decision that it was unlawful in the US, the more angry I become. It just seems to be the product of a society that sees library lending in general as an impermissible market harm.

Particularly galling is the fact that the court saw a change of medium as creating a derivative work rather than a reproduction. Frustrating.

I think the Internet Archive did not do itself any favours by abandoning the 'one copy' principle with the National Emergency Library thing. That was just asking for trouble.

The challenges I am facing with this peer review: the font size on the pdf is too small and can't be changed and I have terrible tremors at the moment. So holding the book still in twitching hands and in line with my twitching head is difficult and I lose my place. The mechanics of reading.

Haemodialysis is 3-4 hours in hospital 3x per week. But it takes 7 hours...

I left at 5pm.

Arrived at 6pm.

Did not start until 7.30pm due to delays on the ward. One earlier patient running late throws the whole schedule.

Finished at 10.45pm.

Waited 30 mins for transport.

Home at midnight.

An interesting response here from Pangram; it turns out their system did not have peripheral knowledge of the Granta essays that were fed into it, so it could not have scored the essay as non-AI based on publication date...

Started 2x 10min tasks on the computer at 9.30am.

Just finished them.

It was like the world was taunting me today with my computers/servers. Anything that could go wrong, did. Some problems were, terribly, my own fault. Others, were not.

I dug out a humorous photo from 22 years ago of me riding a horse. I do not really know how to ride a horse. I also look like I am flying down the final furlong here, with my posture, even though we were actually standing still. I think she was called Sophie.

Me, 22 years ago, on top of a white horse - me leaning forwards

Life expectancy on haemodialysis is 5-10 years. I am creeping towards that lower bound.

But these figures are also skewed by an older population. So I should/might have a bit longer.

My consultant optimistically gave me 20 years. This can happen. But given my co-morbidities, it's less likely.

I will stop talking about health things, soon. But not now.

Today is my 3rd anniversary of starting haemodialysis. I will "celebrate" by... going to haemodialysis 😂🎉

An unrelenting week.

Skin graft surgery, then dental exam, then IV feeding & haemodialysis, then the new IDMS install at Knowledge Commons, then more IV feeding & haemodialysis, then tooth extraction. Tomorrow is IV feeding & haemodialysis. Then, the day after, just haemodialysis.

Then I turn 40.

Lots of kind people remarking on the string of things I'm going through, medically, right now. It's true. Thank you.

But it's also just ONE single thing: immunodeficiency caused by rheumatoid arthritis.

People don't realise how bad the consequences of this can be!

Research Professional on the stakeholder survey conducted by the Office for Students: "When it comes to higher education’s bigger players, it would seem the OfS is as popular as a rattlesnake in a lucky dip."

This morning, I will have my back left top molar tooth extracted because it has an infected pocket below it. Great.

But a note: it's all a result of rheumatoid arthritis and its treatment. I have strong immunosuppression for it. This let BK virus destroy my kidneys, meaning dialysis for life. It has also given me the chronic norovirus (& malnutrition), a sinus infection, a chest infection, and a tooth infection.

Is it not very likely that Pangram's training set (the AI detector is itself an AI) included the prize's 15-year history in its non-AI training set, so it would rule them out?

That said, my favourite line is just that this is straightforward bad writing.

Introduce yourself with 10 bands/musicians you've seen live.

Fluke
Leftfield
The Chemical Brothers
Rammstein
Rage Against the Machine
The Prodigy
Fat Freddy's Drop
Richard Thompson
Mark Knopfler
Jake Wilson

Just found out that what I am getting is called Intradialytic Parenteral Nutrition (IDPN): an intravenous infusion of essential nutrients—including proteins (amino acids), carbohydrates (dextrose), and fats (lipids)—administered directly into the bloodstream during a hemodialysis session

We're doing the major @kcommons.bsky.social login-system upgrade today. Eeek! A huge piece of work that we are now lifting into place.

A percentage indicator showing 1hr 55mins to go.

Very much looking forward to the Textual Ecologies conference!I will be speaking on a panel with @mkirschenbaum.bsky.social and Elaine Treharne :)

Textual Ecologies: Media, Materials, Environments 15–17 July 2026, University of Passau Speakers: Tania Aguila-Way Edward Allen Miranda Burgess Luisa Calè Alex da Costa Polly Dickson Caroline Edwards Martin Eve Martin Foys Amy Gore Matt Kerr Matthew Kirschenbaum Karis Kraus Victoria Mills Hannah Ryley Susanna Sacks Devani Singh Tim Shao-Hung Teng Elaine Treharne Conference Organizers: Katharina Boehm, Luisa Calè, Orietta Da Rold, and Victoria Mills

The problem described here seems to be that the AI provided more information to doordash drivers who were then able to optimise their routes and timings, at the expense of the first pizzas cooked being delivered later and amalgamated with later pizzas.

Emergency dentist this morning because I definitely have a sinus infection and want to check it wasn't of dental origin. Tedious and expensive!

Today, it's pre-flight checks on the new IDMS for Knowledge Commons, which we are installing tomorrow. I'm very nervous about this. Once we begin and take the site down, the clock starts ticking.

First, though, I must go to the dentist. Also: wound hurts quite a lot today after surgery.

My 40th birthday present looks likely to be... Hard drives, at great expense.

The price of large HDDs is truly staggering. It is, as @gbilder.com put it to me, an assault on personally owned machines, via the AI mania.

Scott is, of course, right: it is possible in compartmentalized spaces for sub-components to be understood purely by specialists. And yes, more specific attribution of function via CRediT is the best answer, so you aren't putting your name to things you don't understand.

Today, I am in hospital having skin graft surgery on my dog bite wound.

Anaesthetist thought I was complicated due to medical background and discussed in detail.

The surgeon thought it looked quite simple. "Take it from the thigh, whack it on the wound" was the phrase she used, I believe.

The protestations here are ridiculous. Even in co-authored scientific pieces spanning multiple fields, every author should have read and understood every cited article. You should not put your name to something that you don't understand.

What could constitute a "fake doi"?

DOIs are not actually links, but identifiers, resolved using the Handle System. Current Crossref guidelines have them shown as links as this is the most common use case.

But, like a person's name, there are a few ways we could speak of fake names or fake DOIs.

The ontology I can think of:

* A real or fake article cited using another article's real DOI
* A fake article's metadata registered as a real DOI (can resolve anywhere)
* A DOI-like URL that does not actually hyperlink to a registered DOI proxy

* An unregistered DOI-like URL that does hyperlink to a DOI proxy (404)
* A DOI-like URL that contains prohibited characters (may not resolve)

Now it's time to rebuild the RAID array. This'll be fun.

I really do find Synology NAS boxes pretty much a joy to use. They just work, a lot of the time. Nice web UI for day to day config. Then I can SSH in and do all sorts of fun if I want. Keep getting tempted by UGREEN but come back to Synology.

A stack of hard drives ready to load! The countdown/installer screen of the DS1819+ migration process.

Today, I have mostly been catching up on sleep after late night dialysis yesterday and then being up half the night. However, I did repair the power supply unit on this Synology NAS. And it works! So that's a huge plus...

The inside of a Synology 1819+ NAS unit, with the power supply unit visible

As an idea of how run-down the dialysis ward is here in the hospital: I have been coming back here for 3 sessions.

In 2 of those sessions, I have just been about to start and the machine has broken down.

Today, 2 machines have broken down on me in a row. Supposed to start at 6pm. It's now 7.40pm.

You would think that even the most pro-AI folks would see this and think: "this is the most terrible idea I can think of".

Unfortunately, they will not. They will see this as the way of the future. Handing over all your financial data to billionaires.

Unbelievable.

Something I learned this morning: in 1931, someone tried to introduce the term "booksneak" to denote the person who came up with the idea of libraries being able to lend books rather than them just being sold. Obviously, it didn't stick!

Things in the world of immunodeficiency continue to go well. One thing after another.

I suspect I have a sinus infection caused by a tooth. I can tell this because every time I blow my nose or bend over, I get a totally foul smell in my sinuses that is not (I am assured) in the environment.

Great.

If Bluesky had a content warning option for images that didn't display them by default, I'd show you the dog bite wound. It is way too gruesome to post without.

Here it is on Mastodon: https://hcommons.social/@mpe/116583316350618799

It won't show unless you click it.

This is what I am going through with the wound. Skin graft surgery on Monday.

Meanwhile, I am hunkering down to a peer review. I have had to miss the deadline on this due to my recent hospitalisation for the dog bite and then also due to the IV feeding fiasco. I try never to miss deadlines on these as I know authors are waiting, but this could not be helped.

As someone who has had chronic norovirus since January and now needs IV feeding as a result, trust me: it isn't "mild".

This piece does bug me.

“This is anti-rigor” is the argument that I was having with it, throughout. And I don't accept the given argument that "you don't evaluate all software you use".

You are being asked to evaluate this thing you made. Without coding knowledge. How?

https://medium.com/@jurgens_24580/everything-we-teach-is-right-but-all-our-classes-are-wrong-94e43f569cde

Is it people as glorified unit tests? Given a range of inputs, does the software give the correct answer, as intended?

Yes, lots of compelling arguments here, but also a proposal for a MAJOR structural change to an education programme.

I also lost my lovely academic office in London in favour of bleak "co-working, hot-desking spaces".

Although, tbf, the building was in such a state of disrepair/age that my office was not wheelchair accessible. So I'd already lost it by the time I needed the chair.

A photo and situation from Mastodon that I can't get out of my head. Just so appalling.

“The Sergeant at Arms blocks Representative Justin Pearson from Memphis from entering a committee meeting about redrawing the map specifically for the district he represents.” - Brian Garlick

“The Sergeant at Arms blocks Representative Justin Pearson from Memphis from entering a committee meeting about redrawing the map specifically for the district he represents. A white officer with a badge, blocking a black congressional member, from joining meeting that specifically targets the black district meant to give representation to people of color. The segregation is alive and well in the south.” - Brian Garlick

I am employed through a UK company that handles international employment for US-based organizations.

They just sent a questionnaire asking whether we would like to be paid in crypto. 🥴

OK, so it was a regulated stablecoin (USDC) but still: wow and what?!

Perhaps I should ask to be paid in Monero, the completely anonymous crypto coin that seems prevalent on the dark web. 😂

The real reason, by the way, will be them avoiding currency transfer fees. My guess.

If you think your Bluesky notifications channel is too quiet, I recommend posting something fairly neutral about AI. You will soon have a diverse set of voices in your channel, cleanly divided into absolute haters and enthusiastic supporters. 😂

Surely it actually shows that the public are willing to elect people from these demographics to represent them?

Yesterday was also a remarkable day in some senses. We had the animal behaviourist come round to work with Luka again.

After this session, I had Luka eating out the palm of my hand in the sitting room, where he is normally totally terrified of me and where he bit me five weeks ago.

Normally in the sitting room, Luka wants to be as far away from me as possible and often won't come and sit even near Helen if I am there.

Yesterday, not only did he take food from me (which is unheard of) but he came & slept right next to my chair. He was putting his nose up on the chair itself.

It may all be cupboard love, but I don't care. I can convert that into real love. Just also, the association of positive food experiences with being near me seems really crucial.

It is almost like a joke, though, that Nikki turns up and Luka is the best-behaved dog on the planet and comes near me and takes treats from me. When she's not there, my word, it is so much harder work.

The Typeless keyboard, which is a system for dictating rather than typing on your phone, has finally done the thing that is needed and added a full keyboard if you swipe across. So you can now type a message with your voice and then edit it with a proper keyboard. This actually works really well!

Of course, you don't always want to type with your voice. In public places, such as trains, it would be hugely awkward and antisocial. However, I am so shaky that it is incredibly useful to be able to just type with my voice.This whole message was composed with Typeless without any correction.

Gearing up for skin graft surgery on Monday & first IV feeding/dialysis hybrid session in hospital, this evening. Meanwhile, rheumatology have badly messed up my drugs and it's a nightmare.

So I have treated myself to a day off work. I am going to the local wildlife park. Gonna see some capybaras!

Will be the first outing in the folding powerchair. Really hope it does its job!

Trying not to feel downhearted about being dragged into hospital for the feeding/dialysis. But I have lost the freedom to do dialysis around my life and, once more, have to organise my life around dialysis.

People don't quite get this. Some surgical coordinator: "ah, so if you have dialysis then, that day is a bit tricky?"

Me: "it's not 'a bit tricky', it's totally impossible. Dialysis is immovable, and if I don't have it, I die."

Anyway, chin up, carry on. That's all there is to do.

Going back to the animal park, you can expect photos.

Down memory lane from when we had Toby, a fair few years ago. He was a grumpy old git, but we loved him so much.

A picture of me studying (using a tablet device with a stylus) sitting next to my Jack Russell terrier, Toby

I have submitted a book chapter!

Now: haemodialysis. Last session wholly on home dialysis, before I must head to hospital 3x per week, for dialysis and IV feeding.

No flexibility now about when to dialyse this way and it's basically 6 hours every other evening, gone. But it will keep me alive.

“You taught me language, and my profit on't / Is I know how to curse” (The Tempest, act 1.2.363–64)

Anthropic claims that Claude's behaviours are getting worse because people wrote of the theoretical harms of AI and now it is trained on that material.

I really like this. I still have an urge to refine it, to say that such a "product" is somewhere in the synthesis of a researcher and their research (project). But it's about a human understanding something new and showing that they do. All about the person.

One annoyance with Zotero (I voice this reluctantly since Zotero is probably the most amazing piece of software I use): when you switch from footnotes to author-date in the word processor (because you didn't read the style guide properly 🙄), you have to reformat the reference and text. Darnit.

A fascinating read. I think @wtgowers.bsky.social might underestimate how totally incomprehensible much of this blog post is to someone with no mathematical training. This is an understatement!: "people who have...
https://gowers.wordpress.com/2026/05/08/a-recent-experience-with-chatgpt-5-5-pro/

... themselves solved difficult problems are likely to be significantly better at using solving problems with the help of AI".

But also: the goal of any knowledge form should be for people to know it. Not just machines. This is why we have PhD vivas. If a machine can propose something and then...

... You can understand it and verify it... And persuade a viva jury that you understand it, this is some form of knowledge gain? Unsure.

Anyway, the blog post was great and very interesting. The bits that weren't written in some alien script, that is.

Also "with no serious mathematical input from me" is so amusing, given that I could not even BEGIN to ask the questions here, let alone understand the mathematical notation (and reasoning) needed to frame them.

I got gasps when I said "eugenics" on a well-known podcast. It was clear I'd crossed a boundary of decorum or was thought of as a conspiracy theorist.

But it was basically UK government policy. "It only kills people who are weak and sick, so they must stay home while everyone else goes about life".

"Or risk their necks! Hey! That's their choice now!"

Anthropic on X: "We started by investigating why Claude chose to blackmail. We believe the original source of the behavior was internet text that portrays AI as evil and interested in self-preservation."

So we wrote that AI might be evil and then they trained the AI on that text, so it was evil.

Original post: https://x.com/AnthropicAI/status/2052808791301697563

This morning's research was on Gothic Marxism, via Cronin, James, and Sophie James, ‘Consumer Culture Gothica: Marx’s Phantasmaterialist Influence on Critical Consumer Research’, Consumption Markets & Culture, 28.6 (2025), pp. 472–89, doi:10.1080/10253866.2025.2497813 and

Cohen, Margaret, Profane Illumination Walter Benjamin and the Paris of Surrealist Revolution (University of California Press, 1995).

Enjoyed this description: “Marx does not disqualify supernatural language and signifiers as unscientific delusions...

... rather he encourages a broadening of materialist analysis to better consider how and why the irrational and absurd, as represented by ghosts and phantasms, penetrate market societies”.

This is, of course, a very widespread view, just here summarizing Derrida and Cohen.

But it prompted me to consider the irrational & absurd aspects of AI/LLMs and how these aspects, too, "penetrate market societies".

(This is for a piece on AI and the Gothic)

Happy 89th birthday to Thomas Pynchon!

I won't be doing any "Pynchon in public" as I'm too unwell but good luck to those who do!

I turn 40 this month. I am struggling to think of anything I can do to celebrate. I have to go for dialysis and IV feeding in the late afternoon/evening. I am not very mobile at all at the moment, but now have a fold-up powerchair. We live in Kent on the South-East coast. Any ideas?

It is hard to explain how much time in my life I spend just organising medical admin. When you are chronically unwell, you can spend a huge portion of your time doing this. And it's all tripped up by things like this: Why can't I cancel online? What do you mean?

A picture saying: 'This appointment cannot be cancelled online. To cancel, please call' (followed by a phone number).

Gray, Anna, et al., ‘Understanding the Psychological Impacts of Shielding From COVID-19 for Immunocompromised Individuals’, Health Expectations, 29.2 (2026), p. e70675, https://doi.org/10.1111/hex.70675

With my immunodeficiency, shielding is still here. We must weigh benefit vs danger in each encounter.

We use a system called HelpScout at KC that handles user queries. This reminded me of my teenage years as a Scout.

So here's a picture of me, age 18, as a Scout, having just climbed one of the world's highest active volcanos, Cotopaxi, in Ecuador, with my longstanding friend, Julian.

Two men standing on an icy summit with a large volcanic crater behind them. They are holding ice axes. They are holding each other's hands in a locked formation in celebration of reaching the top. I (Martin Eve) am on the right, smiling triumphantly. The other man, Julian Cottee, is wearing a bobble hat.

In other news, I forgot to mention yesterday that my new power chair has a remote control. Someone else can drive! This can be really useful for people who can't steer the wheelchair themselves. In my case, it just provides entertainment for Mrs Eve to joke about where she's going to send me!

Struggling a bit this morning, in one sense. My hands are extremely tremor-ey/shakey. Mouse keeps jumping as I twitch/tic. Keyboard not easy to use. I keep actually dropping/throwing my phone if I use that! Frustrating.

I have a new, folding powerchair! This will fit in the car so we can now go places and DO THINGS! Helen Eve can fold it and lift with ease. Hard to say concisely what this means, but the nearest term is: "freedom". An example: yesterday was a bank holiday. We stayed at home. We thought ... 🧵1/3

... it would be nice to go to a wildlife park, for instance. But I cannot walk any distance when we get there. Pushing a wheelchair is really hard work, so that didn't appeal much to either of us. This solves that problem! It's nowhere near as robust as my Q500. That thing is a tank that ... 🧵2/3

... can climb curbs and speed about, with a dead-stop handbrake. This is slower, less powerful (can't climb curbs), and with gentle braking. So no speeding about. But it will do the job. Get somewhere, be able to move about, have a bit more of a life. Hurrah! 🧵3/3

The Motion Photon Lite powerchair. An electric wheelchair.

I think this is the first citation of one of my Bluesky posts (by @jefrantsvag.bsky.social)

Frantsvåg, Jan Erik, ‘Reflections on the Demise of First Monday’, Nordic Perspectives on Open Science, 11 (2026), https://doi.org/10.7557/11.8673

A sad topic: the death of the long-running OA journal, First Monday.

I am sick to the back teeth of the vile, yet soon-to-be democratically sanctioned, culture in the UK.

In fact, sick of the already sanctioned government culture. Austerity, kick the disabled on benefits, destroy universities, privatise the NHS.

I want to get off this ride.

This recent US import of punishing other districts for voting for alternative representatives is just sickening. But wait: the concept of detention centres for "illegal immigrants" is also horrifying and based on ICE concentration camps.

Again: please stop this ride!

"Trump says US navy will ‘guide’ trapped ships from strait of Hormuz amid ‘very positive’ talks with Iran"

Two minutes later: "Iran’s military is reportedly saying US forces will be attacked if they enter the strait of Hormuz".

Nothing Trump says can be taken as remotely true.

Hoping that I will hear about my IV feeding schedule this week. Have been priming my laptop ready to work while on the ward.

I have hacked together a neat little tool. It's called opfs https://github.com/MartinPaulEve/opfs. Basically, some things that we do as developers require files on disk, that have credentials in them. AWS is a bad example of this as are fstab credential files for CIFS. Another example ... 🧵1/3

... is AI token credential files. These are unsecured secrets that just exist in the open on our machines. I wrote a tool that virtualizes those files. So ~/.opfs/nas1 appears on the disk, but it's actually simulated. What happens when you try to access that file is that it looks up the ... 🧵2/3

... secret that is associated with it in 1Password. It then asks the user to confirm access. Then the file contains the 1Password secret! The project uses the excellent FUSE library and is Linux-only at the moment, but it should be portable to Mac, given that there is FUSE for Mac. 🧵3/3

I have had a NAS unit die on me. Most likely: power supply needs replacing. However, I needed to get access to the data and cannot replace the PSU at the moment. So: I transferred the disks to another 8-bay NAS. It recognised 7 of the 8 drives and, with 2-drive failure tolerance, I have the data!

The reason I cannot replace the power supply is that I am too weak at present to move the things needed to reach the NAS and then to lift it, even empty of disks.

I have now successfully recovered the data, repairing the RAID array using the fault tolerance. Feels like a blooming miracle!

Totally true.

But a policy or bounty doesn't even mean that inquiries will be welcome. I showed a full admin takeover to a startup with a bug bounty, just under two decades ago. Full working proof of concept, but I had to argue so hard for payment. They eventually paid, but wow was it difficult.

Had a panic last night. Tried to login to my Zotero account only to find that there was new 2fa - by email. However, I had not updated my email from the old @martineve.com address. I added the new https://eve.gd address from a logged-in account. But this shows a weak design pattern...

When a new address is added, confirmation should really go to the OLD address. If you just confirm the new address, you could just be confirming an attacker's controlled email. But it did let me recover my account 😀

I turn 40 this month. I never thought that the most seemingly exciting "present" that I might covet at this age would be a folding powerchair that can fit in our car. But it is, because it would let me go places in the car and be able to move at the other end. I want this one.

That said, there are some absolutely terrible ones that look like a Lego kit, quite frankly.

The Sunrise Medical Quickie Q50 R Carbon Comfort Folding Powerchair. It is an electric wheelchair. The Sunrise Medical Quickie Q50 R Carbon Comfort Folding Powerchair folded up into a neat upright package. A mobility scooter that looks extremely uncomfortable and flimsy.

"we call out these violations of the fundamental freedoms of peaceful assembly and association, and interference with the freedom of expression and civic space of the entire RightsCon community"

We did our test IDMS deployment on Wednesday. Momentous. It was almost a total success. Almost. I needed 15 more minutes this morning (I was off yesterday) to get it working.

I hope to publish a "postmortem" blog post about it all later today.