The UK Mega Dinghy

It looks like the UK is getting a new national humiliation each week lately: The migrant mega dinghy is a national humiliation. Well, if yesterday's post here is any indicator, if your ruling class has people with names like Sebastian and Zanny in it, you deserve every national humiliation you get. (Full disclosure: there is a squirrel in our back yard that hadn't been named until yesterday. Zanny it was.)

It looks like Andy Burnham has achieved what his predecessors could not. He has found a solution to the small-boats crisis. . .

In the early hours of Monday morning, a 49-foot dinghy carrying 230 illegal immigrants arrived on British shores from France. It is by some margin the largest number of migrants to have arrived in a single vessel. It surpasses by far the previous record of 165, a feat of human trafficking achieved late last month. As many people have already pointed out, the ‘small boats crisis’ may now technically be over, as we enter the era of migrant cruise-liners.

The UK Guardian has more detail on the new trend:

Last year, 106 people were reported to have crossed the Channel in a single dinghy in August and 125 in September. The emergence of mega-dinghies coincided with the announcement in July 2025 of the UK-France “one in, one out” deal, under which a person arriving in the UK from France by small boat can be returned in exchange for another person being brought legally to the UK to claim asylum.

On 12 July this year, a mega-dinghy carrying 128 people arrived in the UK, followed by another carrying 165 people on 23 July. On 4 August, a dinghy carrying 173 people capsized after the engine caught fire. All those onboard are believed to have been rescued, with 11 sustaining minor injuries and three taken to hospital in Boulogne.

The use of mega-dinghies poses an even greater risk to life than smaller vessels, which have recently carried an average of between 50 and 60 passengers, according to NGOs working with asylum seekers crossing the Channel. When small-boat crossings began in 2018, only a handful of people typically travelled in each dinghy.

Each new national humiliation in the UK leads me to undertake a new effort to understand civics there. The first question that popped up for me was, "Where is the UK's Coast Guard?" The answer is that the UK doesn't have one, at least as most contries, including the US, understand the term:

His Majesty's Coastguard (HM Coastguard or HMCG) is the section of the Maritime and Coastguard Agency responsible, through the Secretary of State for Transport to Parliament, for the initiation and co-ordination of all maritime search and rescue (SAR) within the UK Maritime Search and Rescue Region.

. . . His Majesty's Coastguard is a uniformed service that fulfils six of the nine functions required by the International Maritime Organization (IMO):

Hey, six out of nine functions is pretty good, right, at least for a third-world country? The six are Search and Rescue, Pollution Response. Vessel Traffic Management, Maritime Safety, Accident and Disaster Response, and Maritime Security. What's left out?

The other three IMO functions; Customs/Border Control, Fisheries Control and Law Enforcement, are undertaken by the UK Border Force, Marine Management Organisation and local police forces, respectively. . . . HM Coastguard is a civilian agency, unlike other coastguard services around the world, as the duties traditionally associated with a military coastguard service are spread around the UK Civil Service and British Armed Forces.

However, UK efforts to limit migration via dinghies small or large appear to be feckless. According to the House of Commoms Library:

The number of migrants crossing the English Channel has increased in 2021, with the BBC reporting that 12,600 people have made the journey so far this year.

Recent news reports have confirmed that UK Border Force are being trained to use certain “turn around” tactics to intercept and return migrants crossing the Channel.

The Government has said that any plans will be “safe and legal”, but France and other interested parties have suggested such tactics could violate international law.

However,

the UK Government has decided not to implement its plan to use pushback tactics for small boat crossings.

Instead, it appears that migrants who are detected arriving in small boats are taken to processing centers. At that point, about 99% apply for asylum in the UK and are housed in migrant hotel facilities. Most wind up remaining in the UK, whatever the legal disposition of their cases:

Around 2,600 people who had first arrived in the UK by small boat were returned to another country in 2025. In total, there were 7,500 such returns between 2018 and 2025 – around 4% of all arrivals during this period.

It remains unclear how many of the 193,000 people who arrived by small boat between 2018 and 2025 are liable for removal, or are likely to become so in the future. By the end of 2025, around 48,000 small boat arrivals had been refused asylum at the initial decision stage, though some of these individuals will eventually be granted status on appeal and hence be able to remain in the UK. Another 33,000 had their asylum application withdrawn or administratively decided.

For now, the UK appears to lack the political will to control its borders, which has had the effect of institutionalizing smuggling networks:

Smuggling groups have become larger and increasingly professionalised in recent years, and were able to capitalise on their initial experience as enforcement measures were progressively tightened. There is some evidence that the price of crossing the Channel declined over time, although estimates are uncertain and can vary significantly. According to some sources, prices had fallen from around €14,000 in 2018 to well under €4,000 by 2024. Media investigations indicated prices of around €1,500 in early 2026.

There won't be another national election in the UK until 2029.

AI And God

Yesterday I noted that Glenn Reynolds's fundamental problem is that he never got an education. Certainly, if he ever took Philosophy 101, he must have cheated. One of the basic issues covered in an introductory philosophy course is the "mind-body problem".

The mind-body problem is a philosophical issue that questions the relationship between mental states, such as beliefs and desires, and physical states of the brain and body. It explores whether the mind and body are distinct entities or one interacting component, a debate primarily characterized by dualism and physicalism. Understanding this issue is crucial for fields like psychology, neuroscience, and cognitive science, as it influences how we perceive consciousness and self-awareness.

According to a UK study guide,

Physicalism (sometimes called materialism) is the view that everything that exists is ultimately physical. This includes mental states, which physicalists argue can be fully explained in terms of physical processes.

According to this view, consciousness, thoughts, and emotions are not separate from the body but are instead the product of brain activity.

. . . Unlike physicalism, dualism argues that the mind and body are fundamentally different kinds of things.

The most famous dualist, René Descartes, believed that the mind is a non-physical substance that is completely separate from but interacts with the physical body.

. . . If the mind is non-physical and the body is physical, how do they interact? How can something non-physical cause physical changes (like deciding to raise your hand and then actually doing it)?

Reynolds, because he either never took Philosophy 101 or he cheated, endorses the physicalism of the quack Ray Kurzweil; both believe it's possible to upload consciousness into a computer:

Kurzweil notes that computational capacity alone will not create artificial intelligence. He asserts that the best way to build machine intelligence is to first understand human intelligence. The first step is to image the brain, to peer inside it. Kurzweil claims imaging technologies such as PET and fMRI are increasing exponentially in resolution while he predicts even greater detail will be obtained during the 2020s when it becomes possible to scan the brain from the inside using nanobots.

Hey, right, it's the 2020s, so we've already scanned the brain from the inside using nanobots, right? He goes on,

Beyond reverse engineering the brain in order to understand and emulate it, Kurzweil introduces the idea of "uploading" a specific brain with every mental process intact, to be instantiated on a "suitably powerful computational substrate".

In other words, every mental process is a specific physical state that can be uploaded onto a computer. This is an absurdity. But the AI boondoggle takes this a step farther: we don't need to wait until we can upload Glenn Reynolds's brain into a computer to get consciousness, it's going to pop into existence on an AI data center ab nihilo! This is a version of the neo-Darwinian theory that a certain type of primeval goo came along, and hey presto, life started! But where did math come from? It isn't physical, it can't have just evolved.

Another exponent of this weird AI-will-generate-consciousnmess (even though the brain is entirely physical) school of thought is Sebastian Mallaby, son of Sir Christopher Mallaby, who was Ambassador of the United Kingdom to Germany (1988–1993) and Ambassador of the United Kingdom to France (1993–1996), educated at Eton and Oxford, and husband of Zanny Minton Beddoes, Editor in Chief of The Economist, lately known for trying to debate Elon Musk. Just the guy to address the mind-body problem, woudln't you think? And he does it, in all places, at Foreign Affairs:

[C]ontemplating a technology that seems unfathomable in scope, humans reach for the lexicon that exists to describe mystery: that of religion. Nearly a decade ago, Anthony Levandowski, an early actor in Google’s self-driving car project, started a church called Way of the Future, whose IRS filings state that it is devoted to “the realization, acceptance, and worship of a Godhead based on Artificial Intelligence,” according to Wired. Encountering an early chatbot that appeared eerily sentient, the Google engineer Blake Lemoine wrote, “Who am I to tell God where He can and can’t put souls?”

. . . It was only a matter of time before the Catholic Church waded into this maelstrom. Following industrialization in the late nineteenth century, the rise of mass media in the mid-twentieth, and the birth of the knowledge economy in the late twentieth, the church responded with encyclicals, papal letters laying out how Catholic doctrine applied to each upheaval. Now the Vatican has stepped in again, defending its conception of human dignity and the purpose of creation in the face of the AI revolution. Pope Leo XIV’s Magnifica Humanitas (Magnificent Humanity), the first encyclical of his pontificate, is nothing if not ambitious.

. . . As technologists debate whether AI systems’ possible consciousness warrants a machine version of human rights, the pope insists that humans and machines differ fundamentally. AI systems “merely imitate certain functions of human intelligence,” the encyclical states. “Even when these tools are described as capable of ‘learning,’ . . . it is not the experience of those who allow themselves to be shaped by life and grow over time through choices, mistakes, forgiveness and fidelity.”

. . . Whether the church’s assertions withstand analytic scrutiny is a different question. By postulating that AI systems “merely imitate” human intelligence, Magnifica Humanitas rejects the long-standing claim that imitating intelligence is the same as possessing it. Far from suggesting an absence of intelligence, that theory goes, imitation may be a facet of it.

But now he gets to Philosophy 101, the mind-body problem, and the Chinese Room Experiment, which like the pope's encyclical gets things all wrong:

The best case for the Vatican’s opposing view dates to 1980, when the philosopher John Searle proposed what became known as the “Chinese room argument.” The goal of Searle’s thought experiment was to refute the famous Turing test, which holds that a machine should be judged only by its output. According to the test, if the system’s output appears intelligent to humans, then the system is indeed intelligent. By way of a rebuttal, Searle imagined a lone man in a room receiving messages under the door. The messages are written in Chinese, and the man passes back messages in Chinese, convincing his correspondent that he, too, is fluent in the language. But this persuasion is a con. The man composes his responses with the help of a manual that instructs him on which incoming string of Chinese characters should be answered with which other string of Chinese characters.

But that argument, like the pope, is now out of date:

Searle’s rebuttal was powerful during the years before the modern AI revolution. Machines based their outputs on rules that resembled the instruction manual in Searle’s scenario. Although they begin by ingesting words and treating them as statistics, they emerge from their training with a fine grasp of meaning.

Ah, OK, that settles it. But here's a study guide that explains how large language models work:

The key idea is that an LLM isn't "thinking" or "understanding" in a human sense. It's making a statistical calculation: based on the sequence of words (tokens) it has seen so far, what is the most probable next word?

. . . An LLM is often described as a 'stochastic parrot' that predicts the next token.

Let's go back to the lone man in the Chinese room. As of 1980, he was following specific if-then conditions regarding which Chinese message to pass out in response to whatever Chinese message he received. As of 2026, he's been doing this long enough that he has a good general idea of which Chinese message to pass out, and he's at least reaching for the one he's most used to passing out in response to the next one that comes in. But that doesn't mean he understands Chinese any better in 2026 than he did in 1980.

He still can't read a word of Chinese, even if he's now passing out messages that say, "I'm sorry to hear that", since he himself still doesn't know what the message says. But Sebastian Mallaby -- you can't make up a name like that, any more than you can make up Zanny Minton Beddoes -- hasn't been reading his study guides. He goes on,

Today, modern AI systems, trained on text including novels, social media posts, and human emoting of all kinds, recognize anger, joy, shame, sadness, and anxiety. And they don’t merely recognize emotion. Their behavior can be made consistent with particular emotional states or personalities. By tweaking its settings, researchers can make a model upbeat or negative, honest or sycophantic. Unlike calculators, spreadsheets, or autocomplete systems, large language models can acquire what one might call temperament.

No, in the words of LLM experts, all an LLM does is calculate the statistical likelihood of one set of words appearing in the context of another set of words. If I pass words to it that are statistically related to sadness, it will return equivalent words in its answer, but it has no idea what those words actually mean. It only appears to acquire temperament via large-scale statistical association.

In my post on August 7, I got AI to write a micro fiction piece that looked as though Lydia Davis had written it, and it did a brilliant job looking at the contexts of her subjects, grammar, structure, and vocabulary. In fact, it did such a brilliant job that it unintentionally turned out hilarious satire. AI has no idea what satire is, and it has no idea what contexts can be unintentionally funny. In this case, however sophisticated it may have seemed, it was just a guy in a Chinese room.

But now Sebastian reveals his basic lack of any sort of education:

The human brain is a physical object, composed of biological material that obeys the laws that govern the rest of the universe, computers included. Human brains, like computer brains, work on trickles of electricity; when they form an opinion or conceive a plan, they are processing information. The idea that the gooey mass inside the skull contains some ineffable, nonprogrammable essence—consciousness, spirit, or something else—may be intuitively appealing. But as an analytic matter, the idea is flimsy. And it will become even less convincing as AI continues to advance.

In other words, he's a physicalist. A machine brain is no different from a human or animal brain. If you upload Sebastian Mallaby's brain scan into a computer, you have the full Sebastian Mallaby. If the pope says that's not true, the pope is wrong. In effect, Sebastian is saying there's no moral difference between a human and an AI computer, because soon enough, they'll be the same thing. Contra the pope, who makes "slippery claims about human exceptionalism" everyone is basically a machine, especially if machnines can convince humans they're human. Or something like that, because in the end, I don't think he understands he's arguing against himself.

The final conspicuous gap concerns the debate over the nature of machine intelligence. The extent to which AI algorithms generate the equivalent of human thoughts and feelings is a philosophical question that may never be resolved, but understanding exactly how AI systems work is a practical and urgent goal.

But he's just spent 1600 words insisting it doesn't matter whather AI algorithms generate the equivalent of human thoughts and feelings, because they appear to, and the Chinese Room Experiment is beside the point. Now he's saying we have to go inside the Chinese Room and figure out what the guy is actually doing:

Magnifica Humanitas claims that we cannot know the thinking inside the machines, because there is no thinking in them worth knowing. But the opposite is true. Machines will never join our moral universe, but if humans aspire to control them, we must first understand them.

The best I can offer is this is what you'd expect from someone named Sebastian Mallaby who's married to someone named Zanny Minton Beddoes, neither of whom took Philosophy 101, or if they did, they cheated.

Glenn Reynolds's Main Problem Is He's 25 Years Older, But Not A Day Younger

And for that matter, not an IQ point smarter. He's congratulating himself for 25 years of Instapundit, and he's apparently grateful for Dr Fauci's example of how to do this:

Along the way there were podcasts — the late, lamented Glenn and Helen Show, which I really enjoyed, several years with PJTV, and various columns for TechCentralStation, the Washington Examiner, USA Today, nearly a decade of columns in Popular Mechanics, and now the New York Post.

But always there was the blog. Behind everything there was the blog. Most of my ideas came from the blog, even when they appeared in columns, podcasts or TV shows. Many of the ideas on the blog came from readers who emailed me, or, later, commented on posts. And it’s still here. As I said in my brief anniversary observation on Saturday, I never expected to do this for 25 years, and I can’t promise I’ll still be doing it in 25 more. But I can’t promise that I won’t be, either.

I started following Instapundit in the mid-2000s, and the version of what it was that he gives in this piece is misleading to say the least:

In the early days the blog was more focused on tech and pop culture. I remember linking to an article on who had the best abs at the Video Music Awards — I backed Alicia Keys — and similar stuff. However, some later InstaPundit themes appeared: I had a post titled “Blood and Iron” about donating blood and life expectancy, and there were numerous posts on FBI and Secret Service incompetence and wrongdoing, which are themes that persist to this day.

Oh, by the way, there was a little bit about "life expectancy". That's not the Instapundit I was reading. I asked Chrome AI mode, "When was the last time Glenn Reynolds mentioned Aubrey de Grey?" (This, by the way, is where AI really shines.) It replied,

The last major, documented instance of Glenn Reynolds directly highlighting Aubrey de Grey in his professional writing was in March 2008, when he penned an article titled "The End of Aging?" for Popular Mechanics.

March 2008: Reynolds features de Grey's life-extension theories in his Popular Mechanics column.

March 2006: Reynolds publishes his book An Army of Davids, dedicate an entire chapter ("Live Long—And Prosper!") to a comprehensive text interview with de Grey regarding biological "escape velocity".

December 2004: On his blog Instapundit, Reynolds links to a profile of de Grey, noting that he was "seeing this guy Aubrey de Grey all over the place".

Aubrey de Grey is a quack. It took Reynolds some years to recognize this, and although he's never posted anything to the effect that he was taken in at any point, he seems to have dropped the guy by 2008. As of 2008, this was de Grey's line:

To encourage people to take his ideas seriously, Aubrey de Grey, the originator of the strategies for engineered negligible senescence, has organised a competition. He is offering a prize for the development of what he calls a Methuselah mouse. There are actually two prizes to be had. One is for longevity, the other for rejuvenation. The prize for longevity can be won by a new strain of mouse - one bred or genetically engineered to live a long time. That for rejuvenation requires treatment to begin when the mice are already in middle age. ... The winner establishes a record that others have to break. At the moment the records for longevity and rejuvenation are five years and almost four in an animal that normally lives for three.

According to Chrome AI mode, this prize has been awarded many times, for example,

Dr. Andrzej Bartke (Longevity Prize): Awarded for breeding a genetically modified mouse that lived for 1,819 days (nearly 5 years), which is the equivalent of roughly 180 human years. His research involved altering a gene that controls responses to growth hormones.

Except that without any special life-extenmsion measures, genetic or otherwise, a mouse lived for nine years at the San Diego Zoo:

Wikipedia's entry on de Grey suggests that almost as soon as he began to achieve a high media profile, thanks in some measure to Glenn Reynolds, he began to be debunked:

In 2005, MIT Technology Review, in cooperation with the Methuselah Foundation, announced a US$20,000 prize for any molecular biologist who could demonstrate that SENS was "so wrong that it is unworthy of learned debate." . . . The judges concluded that none of the challengers had disproved SENS, but the magazine opined that Preston Estep's submission was particularly eloquent and well written, and awarded the contestant US$10,000. The judges also noted "the proponents of SENS have not made a compelling case for SENS", and wrote that many of its proposals could not be verified with the current level of scientific knowledge and technology, concluding that "SENS does not compel the assent of many knowledgeable scientists; but neither is it demonstrably wrong."

. . . A 2005 article about SENS published in the viewpoint section of EMBO Reports by 28 scientists concluded that none of de Grey's hypotheses "have ever been shown to extend the lifespan of any organism, let alone humans".

According to the Wikipedia entry, de Grey has a contract to have his head frozen at the time of his putative "death", on the assumption that future technology will be able to cure what ailed him and reattach it to a new body, or maybe just upload his brain data to a giant computer or something, the same thing Glenn Reynolds believes. Nor does Reynolds mention his enthusiastic support of another quack, Ray Kurzweil. I asked Chrome AI mode, "Was Glenn Reynolds a major promoter of Ray Kurzweil?" It answered,

Yes, Glenn Reynolds was a major early promoter of Ray Kurzweil and his techno-futuristic visions.

Through his highly influential blog Instapundit, as well as his mainstream media columns, Reynolds served as a crucial bridge between the niche transhumanist subculture and the broader public. He consistently amplified Kurzweil's ideas regarding exponential technological growth, nanotechnology, and the coming artificial intelligence revolution.

Key Ways Reynolds Promoted Kurzweil

High-Profile Book Reviews: When Kurzweil released his seminal book The Singularity Is Near in 2005, Reynolds wrote a prominent, glowing review for The Wall Street Journal titled "Here it Comes: Technology's progress will soon accelerate — exponentially. You have no idea how much. Ray Kurzweil does." He revisited this advocacy by promoting the 2024 sequel, The Singularity Is Nearer.

The "Instapundit" Effect: During the peak influence of Instapundit in the early-to-mid 2000s, Reynolds frequently blogged about "The Singularity," regularly pointing his massive audience toward Kurzweil's essays, interviews, and predictions.

Integrating Kurzweil into His Own Philosophy: In his 2006 book, An Army of Davids, Reynolds relied heavily on Kurzweil's projections of decentralized, democratized technology empowering individual citizens to take down massive corporate and government hierarchies.

I asked Chrome AI mode, "Is there any similarity betwqeen belief in the Singularity and belief in the Great Pumpkin?" It answered,

Yes, there are striking psychological and structural similarities between belief in the technological singularity and belief in Linus’s Great Pumpkin from Peanuts. While one stems from computer science and the other from a comic strip, both function as modern myths.

The Chosen Prophet: Ray Kurzweil acts as the singularity's chief prophet, tracking its arrival with charts. Linus van Pelt is the sole, unwavering prophet of the Great Pumpkin.

The Promised Date: Singularity believers often pinpoint a specific timeframe (like 2045) for the transformation. Linus waits every year on a specific night: Halloween.

A Benevolent Savior: The singularity is often predicted to solve all human problems, cure diseases, and grant immortality. The Great Pumpkin is expected to fly through the air and deliver toys to good children.

Reynolds is a transhumanist who still believes in the longevity tenets of both Aubrey de Grey and Ray Kurzweil, although at this point, he minimizes this. His problem is he's 25 years older and not a day closer to immortality. I won't even bother to touch on his libertarianism. I'm not sure how he differs in principle from Jason Arday; he holds an endowed chair at a major university, but he shows no sign of every having had an education himself.

Politically Incorrect AI

At The Nerve News: Cory Doctorow: The people who tell you ‘AI is changing everything’ are lying

If Donald Trump ordered Big Tech to turn off all of your country's chatbots tomorrow, nothing would change. Every one of your country's ministries and corporations would chug on with nary a hitch. Households, too, though perhaps a few of the younger members of those families would have to do their own homework again.

Contrast this with what would transpire if Trump directed his tech giants to switch off your country's Office 365 access, or to brick your Android and iOS phones, or to killswitch your John Deere tractors. Your country would effectively cease to exist.

Having made a career in tech, I can vouch for this insight:

One person who's had a lot of opportunity to observe the shear between the stated business/AI situation and the real business AI situation is Nikhil Suresh from Hermit Tech, a consulting firm of "radically ethical data wizards" (that is, tech consultants). For a year and a half he has been talking to hundreds of executives — and, more importantly, their subordinates — about what, if anything, AI is doing for their businesses.

. . . Suresh says he's never seen a successful enterprise AI project: "Every single one – we have seen 0% success in a year and a half." Not one of their clients would face a business challenge if OpenAI went out of business tomorrow. The problem most companies struggle with is that they're "terminally bad at running software projects effectively". Adding AI to the mix doesn't solve this problem – it just adds a whole new range of ways that software deployment can fail.

Chatbots don't help. The internally facing chatbot that's supposed to help employees figure out how to navigate the business sucks because it is only as good as its training data – the business's documentation of its own processes. Businesses suck at documenting their processes. Customer-facing chatbots also suck. They either can't solve your problem, or, when they seem to solve your problem, the "solution" goes nowhere.

Suresh recounts his sole positive customer-service chatbot experience: a Mitsubishi chatbot with a natural-sounding, responsive voice politely took all the details of an automotive failure and promised him a callback. That callback never came, but Suresh is certain that Mitsubishi has logged this as a chatbot success story, even though the experience convinced him not to buy a Mitsubishi car.

One thing I'm noticing is that the unquestioning corporate adoption of AI is generating a peculiar inverse reaction in academic and literary environments, for instance at The Yale Review:

At The Yale Review, we are committed to publishing work shaped by human judgment, care, and responsibility. Generative artificial intelligence (AI) technologies—including large language models (LLMs)—are increasingly part of the cultural landscape, both as a subject of inquiry and as a tool in creative and critical work.

As an institution, we may occasionally use AI technologies to support internal tasks such as calendaring, research collation, and editorial planning in ways that we deem in keeping with our overall mission. However, all published writing is selected, edited, and fact-checked by human editors.

We welcome submissions that engage thoughtfully or critically with AI—especially when AI is central to the inquiry or deployed self-reflexively as part of a writer’s process. All work we publish must reflect the discernment and intent of a human author. Where AI-generated material appears prominently in a submission, we expect writers to note this in brief and explain how it was used.

More frequently, the policy is just plain No AI, as at the equally prestigious Baltimore Review:

About AI: Do we really have to say it? OK. No. We want to read work created entirely by humans.

What's puzzling here is that, as we see just above, AI is struggling to prove itself in business environments, but results from the real world indicate that the one field where it's been a roaring success has been academics, where it's turned out to be a whole new way to beat the system. It'll knock essay questions and comp assignments out of the park, and I strongly suspect that even the most respected journals are reluctant to feed recent years' issues through an AI detector. In fact, I'll bet they're terrified of what would happen if they announced a policy of henceforth doing this with all new submissions.

I've been chuckling for two days over the result AI gave me in yesterday's post, when I asked it simply to generate a story that looked like Lydia Davis, one of the most respected current authors, had written it. It was a perect example of what an AI large language model can do. There was plenty of training data -- Davis's stories are all over the web. The large language model was capable of breaking out the highly visible elements of her narratives and style. The result was unintentional, but hilariously accurate, satire.

What do you think The Yale Review would do if I submitted an essay outlining the insights to be gained from yesterday's post? After all, they "welcome submissions that engage thoughtfully or critically with AI—especially when AI is central to the inquiry or deployed self-reflexively as part of a writer’s process." That might be a bridge too far, but I'm not going to rule this out, at least for now.

More Comes Out On The Idaho College Murders

There have been several recent TV documentaries on the November 13, 2022 murders of four University of Idaho students at an off-campus residence. I watched the one on Investigation Discovery, The Idaho College Murders. Quite a bit of the information in the program challenges the established narrative that these were four naive, unwitting victims.

The first thing that struck both my wife and me was the luxury features of that off-campus residence. We'd both attended undergraduate colleges aimed at the upper middle class in the 1960s, and by comparison, our accommodations back in the day were Spartan. Not only did this place have a fully up-to-date kitchen, it had hardwood floors, fancy bathrooms, and walk-in closets. Everything had new paint. Bedrooms were up to the standard these kids could expect at home in upscale suburbs. I can only surmise that college costs are rising in part due to the increased standard of living students are led to expect.

The second thing uncovered in the documentary was that this residence was widely known as a "party house" and something of an eztension of the on-campus Greek scene. The door locks were combination-style; the combination was widely known and never changed. Outsiders were routinely in and out of the place, it simply wasn't secure.

The third thing was that the murders took place about 4:00 AM, while some of the residents were still awake and partying. Victim Xana Kernodle was still awake, having just received a DoorDash order. Two other surviving residents were also awake. The other three victims, however, were asleep.

The fourth thing, which I frankly found astonishing but shouldn't have, was the level of alcohol and drugs in the victims' blood. Chrome AI mode summarized the medical report:

• Madison Mogen: BAC of 0.282 g/100 mL (an extremely high level approaching alcohol poisoning territory).
• Xana Kernodle: BAC of 0.229 g/100 mL (nearly three times the legal driving limit, with trace levels of amphetamines detected at 12 ng/mL).
• Ethan Chapin: BAC of 0.122 g/100 mL (with a therapeutic or elevated amphetamine/Adderall level of 260 ng/mL).
• Kaylee Goncalves: BAC of 0.107 g/100 mL (reflecting moderate intoxication).

All of them were well above the DUI limit of 0.08 g/100 mL, but three, since they were asleep, hadn't cosumed alcohol for some period before they were killed, which suggests their BAC had previously been even higher. Chrome AI mode concludes,

Legal and medical analysts note that these high levels of intoxication significantly slowed the victims' neurological reaction times, impaired their physical coordination, and heightened their vulnerability to a sudden, middle-of-the-night attack.

I'm also puzzled at Ethan Chapin's apparent use of Adderall, which is used to treat ADHD, which frankly seems incongruous for anyone deemed qualified to attend university. Adderall has a high chance of abuse and misuse and may lead to substance use problems, including addiction.

Adderall may lead to physical dependence after prolonged use, even if taken as directed. Your healthcare provider should check you or your child for the risk of abuse, misuse, and addiction before starting treatment and will monitor you or your child during treatment.

Was this being done for Ethan Chapin?

The documentary pointed out that Bryan Kohberger's motive in carrying out the killings has never been established, and neither have the precise circumstances that led him to select and enter the residence. However, I don't think it's unreasonable to surmise that the residence had a general reputation for lax security, the combination for its locks was apparently well known, and it was generally understood to be a party house. That could easily have attracted Kohberger.

This brings me back again to the experience I had exchanging social-media messages with undergraduates when I was involved in the Dartmouth alumni trustee movement in tbe early 2000s. They objected strenuously to having to swipe keycards to get in and out of dorms, when given the current environment, this is simply a minimum standard of security anywhere. Had this been in place at that off-campus residence, the murders likely wouldn't have taken place. If students resist, or neglect to follow, basic security precautions, they aren't being taught ordinary prudence, which ought to be part of any character development.

It's hard to shake the overall sense of entitlement that surrounds student life in the documentary. Nevertheless, there's also a sense of shallow uniformity in all the victims, and this goes to the standard public reaction: these students were living large in a version of the American dream, what a tragedy that their lives were snuffed out. But they were living recklessly, developing habits of alcohol and drug abuse before they were even of age, and not getting any serious advice from anyone -- if anything, their parents were simply enabling the recklessness and entitlement.

I'm not sure my wife and I would have wanted our grandchidren to wind up in that sort of environment. There's got to be a better alternative.

AI Explains Itself

The Jason Arday and Rahila Khan cases make me think of other potential literary hoaxes, for instance using AI. AI tells me that Lydia Davis, portrait above, is a highly respected author of micro fiction. I asked Chrome AI mode, "Can you write me a micro fiction piece as thoough Lydia Davis had written it?"

The New Arrangement

We agreed to divide the books evenly, but he insisted that the anthology of French poetry belonged to him because his grandfather had given it to him, even though he had never read a single line of it and I had translated three of the poems for a journal that paid me in copies. I let him keep it, of which I was quite proud, until I realized yesterday morning that without its spine on the third shelf, the remaining volumes of biography lean precariously to the left, forcing me to look at them slantwise every time I enter the room, which means he is still disrupting my posture months after moving out.

I asked, "Can you explain how someone might detect this was written by AI?" It answered,

An expert reader or an advanced AI detector might identify this text as machine-generated through several subtle linguistic markers:

Over-engineered stylistic mimicry: The piece hits every stereotypical Lydia Davis trope simultaneously—extreme brevity, hyper-fixation on a mundane domestic object, mathematical score-keeping of a relationship, and a single, winding grammatical sentence. Human writers usually display more stylistic variance, whereas AI tends to generate an idealized "caricature" of an author's voice.

Flawless, low-perplexing syntax: The long sentence flows with perfect grammatical precision despite its complexity. AI rarely makes erratic stylistic choices, punctuation errors, or rhythmic stumbles that naturally occur when a human is drafting micro-fiction.

Predictable thematic resolution: The punchline (the missing book causing a physical lean that mirrors emotional disruption) is highly logical and neat. AI excels at making these clean, conceptually satisfying connections, while human writers often introduce a element of messy, unresolved, or surreal emotional truth.

However, I put the story AI generated using a Lydia Davis model into a free online AI detector, and it certified it as 100% human-written.

Keep in mind that AI does not have an inherent sense of humor.

A Precursor Of The Arday Case

The Jason Arday case brought to mind a somewhat similar kerfuffle from the 1980s, the Rahila Khan case. In 1986, Toby Forward, a white Church of England cleric from Hull, began submitting fiction under the pseudonym Rahila Khan, an Asian Muslim mother of two. Forward's side of the story is this:

It came as quite a shock when I lost Rahila Khan. I’d known her for two years and we were very close. I told my agent that Rahila wanted to write a letter to the Guardian about a witless piece that they had done about her, but the agent was firm. ‘No. No more Rahila Khan. She’s finished.’ I was upset. We’d done a lot together in just two years.

Rahila Khan and I wrote our first story for BBC Radio at the end of August 1985. We didn’t put my name on it. It was rejected, as not quite suitable, but with an encouraging letter saying that she wrote well and should try sending it to the women’s magazines. We were more pleased than disappointed, so we set to work to write another – after all, the producer had asked her to send something else and had said that they wanted things ‘with a genuine “ethnic” background’ because they didn’t get many. I managed to separate the responses. The failure was Rahila Khan’s; the praise was mine, secretly.

We had found a gap in the market and we set about filling it. All the stories from then on were accepted. The producer said she’d love to see Rahila ‘if she was ever in London’; she was unfailingly kind in her criticism of the stories and patient with revisions. She took a lot of time with us. It was like getting a free writing-course. I was very pleased for Rahila and didn’t mind that my name wasn’t on anything. I liked it better that way.

After a while we tried some stories that weren’t about Asian teenagers, but the producer didn’t like them and sent them back saying they weren’t as good as the others. She didn’t even suggest revision, which was odd. We didn’t know whether there wasn’t room for ‘English’ stories, or if they really weren’t much good, or if the producer just expected Asian themes from an Asian and English themes from a white person.

To test it out, I wrote a story with Tom Dale. It was a country-house murder with a twist at the end. Very English. We sent it to the head of Morning Story to get a fresh opinion, but it landed on the same producer’s desk. To our surprise, she liked it, and it was broadcast. By now, it was getting quite difficult for me, writing some stories with Rahila and some with Tom Dale. We did them on different typewriters and with a different lay-out. When Tom Dale made a mistake – like not keeping a copy of a script that needed revision – the producer told him off quite sharply, not at all the way she wrote to Rahila Khan; and he was never invited to London to meet her. Rahila felt a bit guilty about this and sometimes Tom sulked and wouldn’t write any more stories, but he had to in the end if something English grew in his mind that Rahila couldn’t use.

Another version of the story is here:

In 1987, the feminist publishing house Virago published a collection of stories featuring British Asian girls caught in the conflicting demands of home and school. The author of Down the Road, Worlds Away was a 38-year-old mother of two called Rahila Khan, whose fiction had already been aired on Radio 4. “With hard-eyed realism and poignant simplicity,” the book’s blurb read, “Rahila Khan brilliantly gives voice to uneasy youth.” Virago’s remit, as Lennie Goodings puts it in A Bite of the Apple, was “to publish the stories of women’s everyday lives, stories previously not thought worth telling and recording,” and these tales perfectly fitted the bill.

The problem was that Rahila Khan did not exist. The author of Down the Road, Worlds Away was in reality an Anglican vicar called Toby Forward impersonating an Asian woman. He practised this deception, he said, because vicars were considered sitcom characters and he wanted to be taken more seriously. Forward decided “to let Virago know what had happened and to trust to their good sense, good business and basic humanity,” as he wrote in a confession piece in the London Review of Books. Goodings, who recently stepped down as Virago’s boss after 40 years with the firm, devotes two pages of her buoyant memoir to the fiasco.

“Outraged” by the deception, Goodings ordered all remaining copies of Forward’s stories to be pulped. “I now see this was a little extreme,” she writes. “We could probably have laughed it off—and then pulped his book.” The scandal, which created much hilarity in the press, fed the nation’s anti-feminism and belittled the urgency of Virago’s larger project. But it also stimulated genuine discussion. “Aside from thoughts on racism, lies, and feminism,” Goodings recalls, “a conversation and questions about the text emerged.”

“Does it matter who wrote it if a book is good? Our answer: well, it also depends how you are presenting the book. Question: what about the Brontës—they published under false, male names? Answer: they had to. Question: can you tell the difference between male and female writing? Answer: no. Question: is it true that female writers of colour have an easier ride than white men in getting published? Answer: no, look around you. Question: did the Reverend Toby Forward have the last laugh on the feminists? Answer: nobody agrees on that one! Question: is it okay for a vicar to dissemble? Answer: hmmm...”

Well, of course, Jonathan Swift wasn't just a vicar, he was a dean. According to Wikipedia,

In 1708, a cobbler named John Partridge published a popular almanac of astrological predictions. Because Partridge falsely determined the deaths of several church officials, Swift attacked Partridge in Predictions for the Ensuing Year by Isaac Bickerstaff, a parody predicting that Partridge would die on 29 March. Swift followed up with a pamphlet issued on 30 March claiming that Partridge had in fact died, which was widely believed despite Partridge's statements to the contrary.

In other words, Fr Forward acted in the best traditiona of both English Literature and the Church of England, at least as they were back in the day. What both the Rahila Khan and Jason Arday cases suggest is that in more recent times, the UK bourgeoisie -- especially the university-educated media elite -- has suffered a profound loss of confidence, to the point that they must go on hopeful searches for non-whites who prosper just as they do, with the inevitable result that they wind up being hoaxed.

Let's keep in mind that non-white immigration to England was the project of elites who wanted to keep working-class wages down. This got atarted after World War II, when to make socilalized medicine and nationalized transport affordable, the elites brought in the first waves to work in those fields. Somehow, places like Cambridge and the BBC need to promote a myth of high-achieving writers and academics from immigrant communities -- and inevitably, opportunists find gaps in tbe market.

It's hard to avoid thinking the basic narrative is this: yes, we brought in former colonials to keep wages down, but look at what's come of it! People in these communities are moving into elite professions like media and academics! It's been all to the good!

Jason Arday is nothing new.