Words of the Day 2026-05-14

On this day (1796): English country doctor Edward Jenner administers his revolutionary cowpox-based vaccine for smallpox in Berkeley, Gloucestershire.

Hallucinate. From my OED app: “/həˈluːsɪneɪt / verb [no object] 1 experience an apparent sensory perception of something that is not actually present: Ben began hallucinating and having fits. ▪ [with object] experience a hallucination of (something): I don’t care if they’re hallucinating purple snakes [with clause] he starts hallucinating that he is Jesus. 2 Computing (of an artificial intelligence program or tool) produce a response that appears to be accurate or plausible but that contains inaccurate or misleading information: any large language model that’s given an input or a prompt is going to hallucinate [with object] the AI hallucinated entire email correspondences that never took place. ORIGIN mid 17th century (in the sense ‘be deceived, have illusions’): from Latin hallucinat- ‘gone astray in thought’, from the verb hallucinari, from Greek alussein ‘be uneasy or distraught’.” • The OED is not always quick to update, so it’s interesting to see sense 2 there already. With usage examples.

“New psychedelic-like drugs could treat depression without making you trip” [Science Daily]. I’m all for new treatments, but what’s the point of psychelics without psychedelia? “UC Davis researchers created brand-new psychedelic-like compounds by shining UV light on amino acid-based molecules. These compounds activated key serotonin receptors tied to brain plasticity and mental health benefits, but surprisingly did not cause hallucination-like behavior in animal tests. Scientists say the discovery could lead to future treatments for depression, PTSD, and addiction without the intense psychedelic experience.” • For the Richard Dawkins’s of this world, who think believe AI is already conscious, have they given consideration to the idea that AIs might hallucinate because they enjoy it?

“Fraudulent citations, blamed on AI hallucinations, are becoming more common in research papers” [STAT]. “Citations in academic papers are intended to ground research in the work that preceded it, over time creating something of a family tree explaining the roots of ideas, protocols, and studies. But a growing number of these citations lead to dead ends. ‘Fabricated’ citations that do not reference real papers are spreading in the literature, polluting the public record of science, a new study published Thursday in the Lancet shows.1 Tools using generative AI are likely to blame, say the Columbia University researchers who authored the paper. One question that all of us have is: ‘Is AI making science more efficient, helping us do better work, or even the same work, but faster, or is it just creating slop?’ We have all of these papers coming out trying to track the use of [large language models] in science, but none of them really tell us anything about the quality. This is one of the first papers that’s telling us something about the quality of what’s being produced with LLMs, and it’s a signal of slop,” said Misha Teplitskiy, a sociologist of science at the University of Michigan who has studied citation practices and AI use among academics but was not involved with the paper.” • Polluting the public record is what Silicon Valley is all about. For one thing, they’ll sell us the tools to clean up the (very) pollution (they created). For another, they don’t accept “public” as a concept.

“OpenAI claims ChatGPT’s new default model hallucinates way less” [The Verge]. ” OpenAI says its new GPT-5.5 Instant model has ‘significant improvements in factuality across the board.’” Because of course they would. It gets better: “The company claims that, based on ‘internal evaluations [BWA-HA-HA-HA-HA!!!!] GPT-5.5 Instant produced ‘52.5%_{MR SUBLIMINAL Note spurious precision} fewer hallucinated claims’ than its Instant model for GPT-5.3 ‘on high-stakes prompts covering areas like medicine, law, and finance.’” • So if you’re a dull normal, “low stakes” by definition, you’ll get a hallucinatory answer if you ask about your child’s heath? (Health not being “high stakes.”).

“Global Salon: Why Finance Should Govern AI” [Global FInance]. “A year ago, I was hearing a lot of people talking about the human in the loop. Well, it’s the human in the lead. The human has to be the one to lead it, judge it. We’re already seeing how AI gets used in a variety of different industries. Banking is a great example, especially in banking and loan decisions. But we know there could be hallucinations. How do you take it, from a leadership perspective, to be able to say that we have the controls in place to make sure we’re getting the business result? Imagine a sales manager in an organization; they could have a finance person at their side, able to navigate how they’re going to make sales decisions. Whether it’s: Should we be selling in this geography? Should we be selling this product line? Should we eliminate this product line? All of that is going to have this finance partnership.” • Love the opportunism! These finance guys are quick, I’ll give them that!

NOTE

1 From the Lancet: “We present findings from a reference-integrity audit of 2·5 million biomedical papers spanning 3 years, showing that fabricated references are embedded in the peer-reviewed literature at scale, and that the rate of fabrication is accelerating.” And so, throught the process of autocoprophagy, fabricated references will also be embedded in AI training sets, also at scale, and also at an increasing rate. Another way of looking at this is that people halting data center construction are defending science.

Dad Joke of the Day: How do you get 100 doctors into a room that only holds 99?You carry the one.

Discern, from my OED app: “verb [with object] recognize or find out: I can discern no difference between the two policies [with clause] pupils quickly discern what is acceptable to the teacher. ▪ distinguish (someone or something) with difficulty by sight or with the other senses: she could faintly discern the shape of a skull. ORIGIN late Middle English: via Old French from Latin discernere, from dis- ‘apart’ + cernere ‘to separate’.” • Closely allied, then, do distinguish. See Bourdieu, Distinction.

“Philippa Foot” [Stanford Encyclopedia of Philosophy]. “We must, in principle, be able to discern the growth of the fern from a growth on the fern that is due to blight, for example, and such discernment requires a conception of how the fern’s life should progress. Hence, the vital operations that characterize something as an organism are intelligible only against the background of a life form.” • Another way of saying that AI isn’t good with context is saying it lacks discernment, the ability to discern.

“Air Force experimenting with using AI for promotion boards” [Military Times]. “The Air Force has quietly assembled an ‘AI Action Team’ to help leaders think through ethical challenges, develop literacy in the force and identify new applications for the tech, the service’s senior enlisted leader said this month. Among those new uses, CMSAF David Wolfe said, was promotion board screening and ranking…. ‘We don’t really do talent management in the Air Force; we do replacement management,’ he said. ‘And that’s on us, to try to get way better at that.’… Current experiments with automating portions of the Air Force officer promotion boards process aim to find those efficiencies, he said. The service is ‘not letting AI pick [officers], but automating the processes that happen in the background so that when the human looks at it, it’s easy to see, easy to discern and gives us a better chance of making a really good decision as we start to really dive in,’ Wolfe said.” • And an AI’s sycophancy won’t skew what’s “easy to see, easy to discern”? Doubtful. And speaking of the hiring process—

“So you think you want to be ordained…” [Episcopal Diocese of Michigan]. “But how does one get to be a bishop, priest or deacon in the church? The key word is discernment. Discernment is the process by which an individual in the context of Christian community seeks to hear, understand and respond to God’s unique calling. It is really important to remember that discernment is for everyone and that everyone has a calling. As St. Paul writes, ‘Now there are varieties of gifts, but the same Spirit; and there are varieties of services, but the same Lord; and there are varieties of activities, but it is the same God who activates all of them in everyone.” (1 Corinthians 12:4-6)…. It is helpful to think of the entire process as concentric rings of discernment. In the Episcopal Church, the individual’s own sense of call is important, but it must be balanced and supported by the discernment of the whole community. This begins with the person’s priest, but expands out to include the bishop, the congregational discernment committee, the vestry, and diocesan bodies like the Commission on Ministry, the Standing Committee and in some dioceses the Examining Chaplains. All are praying, all are discerning, all are listening for the voice and presence of the Holy Spirit.” • Leaving out the theology, the institutional “concentric rings” structure is interesting. Apparently, neither academic peer review nor Air Force promotion boards employ it.

“The more young people use AI, the more they hate it” [The Verge]. “The fact that so many young people are well aware of these dangers even as they make use of the tools shows that they aren’t buying the hype of AI boosters like OpenAI’s Sam Altman, who has frequently tried to pitch chatbots as tools for doing everything from writing essays to raising a child. Instead, it suggests that Gen Z is hyper-aware of the tools’ limitations — from their well-documented tendency to ‘hallucinate’ made-up information to the social and emotional cognito-hazards of relying on machines for human advice. ‘Altman talks about the technology like it is magic. He has used those words precisely, calling ChatGPT ‘Magic Intelligence in the Cloud,” said [Alex Hanna, the director of research at the Distributed AI Research Institute (DAIR)]. “Gen Z is more realistic about what the tools actually can do. They can handle text-based work that they don’t want to do or feel pressured to do. But they are often rather savvy about their limits.’ This is true even among those who aren’t “anti-AI” and say they find chatbot tools useful. ‘I spend a lot of time thinking about this stuff and I’ve personally come to the conclusion that it’s a load of bullshit for outsourcing jobs,’ Emma Gottlieb, a borderline Zoomer-millennial who works in technical sales for a company that makes equipment for the film industry, told The Verge. Gottlieb says she often uses AI tools to quickly sift through large volumes of technical documents for her job. But she knows better than to take the systems’ outputs at face value. ‘I definitely do double-checks, personally.’” • Assuming for a moment that generatios have agency (“young people” is a lot better, we don’t seem to have a name for what Gen Z is doing here, from finding “limits” to AI, not taking its outputs at “face value,” and looking at is larger effects. “Discernment” seems like a good candidate.

Fortune: “Whenever you find that you are on the side of the majority, it is time to reform.” — Mark Twain

Brain, from my OED app: “brain /breɪn / noun 1 an organ of soft nervous tissue contained in the skull of vertebrates, functioning as the coordinating centre of sensation and intellectual and nervous activity: [as modifier] a brain tumour. ▪ (brains) the substance of an animal’s brain used as food. ▪ informal an electronic device with functions comparable to those of the human brain: an electronic brain. 2 intellectual capacity: I didn’t have enough brains for the sciences [mass noun] success requires brain as well as brawn. ▪ (the brains) informal a clever person who supplies the ideas and plans for a group of people: Tom was the brains of the outfit. ▪ a person’s mind: a tiny alarm bell began to ring in her brain.” Conflating mind and brain. “Brain” sounds like one of those good old monosyllablic Old English_{MR SUBLIMINAL Not the furniture polish!}, so the etymology: “Old English brægen “brain,” from Proto-Germanic *bragnan (source also of Middle Low German bregen, Old Frisian and Dutch brein), a word of uncertain origin, perhaps from PIE root *mregh-m(n)o- “skull, brain” (source also of Greek brekhmos “front part of the skull, top of the head”).” • “Brain” is “a word of uncertain origin”!

One of my favorite punk tunes:

“Deep Belly Fat Linked To Faster Brain Shrinkage” [StudyFInds]. “A large study found that people who reduced and sustained lower levels of deep belly fat showed measurably slower brain shrinkage and better memory and thinking scores years later. General weight loss, measured by BMI, did not show the same brain-protective pattern. It was specifically visceral fat, the kind packed around internal organs, that drove the association with brain health.” • News you can use! Thinking for a moment in the analogic mode of the body politic, I suppose you could conceptualize The Empire/the deep state/The Blob/the Military Industrial Complex as visceral fat.

“Your AI Use Is Breaking My Brain” [404 Media]. “Much has been written about ‘AI psychosis,’ the nonspecific, nonscientific diagnosis given to people who have lost themselves to AI. Less has been said about the cognitive load of what other people’s AI use is doing to the rest of us, and the insidious nature of having to navigate an internet and a world where lazy AI has infiltrated everything. Our brains are now performing untold numbers of calculations per day: Is this AI? Do I care if it’s AI? Why does this sound or look or read so weird? Does this person just write like this? Is this a person at all? I see AI content where I’m conditioned to expect and ignore it: In Google’s ‘AI Overviews’ that famously told us to eat glue pizza, in engagement-bait LinkedIn posts, and throughout our Facebook and Instagram feeds. But increasingly I have the feeling that it’s everywhere, coming from all directions, completely unavoidable. It’s not exactly that I have a revulsion to AI-assisted content or don’t want to get fooled by it. It’s that something is happening where my brain has become the AI police because everything feels incredibly uncanny. I will be going about my day reading, watching, or listening to something and, suddenly, I notice that something is wildly off. Quite simply, I feel like I’m going nuts.” • Indeed. The article goes on to list some “AI tropes.” “Changes everything” is a good one. “Quietly changes everything would be better….