This is Naked Capitalism fundraising week. 525 donors have already invested in our efforts to combat corruption and predatory conduct, particularly in the financial realm. Please join us and participate via our donation page, which shows how to give via check, credit card, debit card, PayPal. Clover, or Wise. Read about why we’re doing this fundraiser, what we’ve accomplished in the last year,, and our current goal, supporting the Coffee Break/Sunday Movie features.
Part the First: Mother Nature Has an Answer for Everything. This first part is a subject I have been following for a long time, since the first research on how plants react when pathogens start munching on them. When this research was begun by several laboratories I knew well, most other scientists in our orbit laughed, or shook their heads. This is a very technical report but it describes just one of the wonders of evolution: Glycan recognition by a plant damage-sensing immune receptor. From the Abstract:
Plants rely on their cell walls not only as structural scaffolds but also as dynamic frontlines where environmental adaptation and immunity start. Composed largely of cellulose, hemicelluloses and pectins polysaccharides, cell walls form the first barrier that pathogens must overcome. To breach this defense, many pathogenic microbes secrete cell wall–degrading enzymes (CWDEs), releasing polysaccharides fragments into the extracellular space. While they can act as nutrients for microbes these wall-derived glycans, serve as DAMPs, signaling danger to the host, and activating its immune defenses. Cell wall-derived DAMPs (Damage-Associated Molecular Patterns) include β-1,4-glucan fragments of cellulose, oligogalacturonides (OGs) from homogalacturonan, xyloglucan- and mannan-derived oligosaccharides, as well as arabinoxylan, xylan and mixed linked-glucans (MLGs)-derived fragments, which trigger immune outputs when applied to plants. Despite their importance, however, the molecular mechanisms by which plants perceive cell wall glycans and integrate these signals into immune pathways remain poorly understood. Here, we address this longstanding question by structurally and functionally characterizing the multidomain receptor kinase Impaired in Glycan Perception 1 (IGP1)/Cellulose Oligosaccharide Receptor Kinase 1 (CORK1), which binds cellulose-derived oligosaccharides. These findings establish a mechanistic framework for DAMP perception and immune signaling activation in plants.
Translated, this means that breakdown products from the infection signal the plant to mount a defense against the pathogen. This reaction can be fairly called an immune response. Part of the downstream responses (e.g., protein kinases) are similar to those in animals when we are infected with bacteria, fungi, worms, or protozoans. Apologies, for getting this far into the weeds, but evolution is a grand thing and sometimes I cannot resist the rabbit hole. Or as Charles Darwin put it at the end of the first edition of The Origin of Species:
There is grandeur in this view of life, with its several powers, having been originally breathed into a few forms or into one; and that, whilst this planet has gone cycling on according to the fixed law of gravity, from so simple a beginning endless forms most beautiful and most wonderful have been, and are being, evolved.
Back in the day I copied this on the first page of my laboratory notebooks. Most of my colleagues just shook their heads. They were not wrong. Nevertheless, a deep understanding of evolution is essential for discovery in biology. Or as Ernst Mayr, another favorite of mine who delightfully took no prisoners during argument, put it: “Our understanding of the world is achieved more effectively by conceptual improvements than by discovery of new facts.” Anyway, this kind of research is most certainly not on the priority list of the Current Administration. Mores the pity.
Part the Second: Brain Science in this Modern Age. The brain is the seat of mind, but that is about as far as brain research has gotten. It is unlikely the trillions of connections that produce “mind” will ever be parsed, but that does not mean that the more scientists know, the better we can understand how the brain works. From a News and Views article in Nature (no archive yet but the links are open access):
A collaborative project from the PsychAD Consortium has compiled RNA-expression (transcriptomic) data from millions of individual brain cells in samples provided by nearly 1,500 people. Writing in Nature, Yang et al., Lee et al. and Venkatesh et al. used this resource to identify a wealth of insights that could shift neuroscientists’ understanding of an area of the brain that shapes human personality and behaviours.
Readers might recall the story of Phineas Gage, a US railway worker who was impaled through the cheek with an iron rod in 1848, destroying the left part of the front of his brain. Despite surviving and returning to an independent life with his memory, physical strength and basic cognition intact, Gage experienced profound personality changes — he was rude to colleagues, rejected advice and became irrational. This revealed, for the first time, that the brain’s frontal lobe is involved in behaviour and executive function, which are the cognitive processes needed for self-control and planning tasks.
Today, neuroscientists know that a part of the frontal lobe called the prefrontal cortex is responsible for executive function. Neural circuits in this region control processes related to attention, working memory, reasoning, planning, decision making, language production and emotional regulation. Compared with those of other mammals, including other primates, the human prefrontal cortex is larger relative to the rest of the brain and exhibits greater connectivity. These features might contribute to humans’ greater working memory, our ability to understand abstract concepts and our capacity to draw connections between ideas.
All true. And one of the more interesting results of this research different patterns in context of brain diseases:
Complementing this work, Lee et al. present a transcriptomic atlas of the dorsolateral prefrontal cortex that is also specific to cell types but placed in the context of disease. The authors made use of the full PsychAD data set, which comprises nearly 1,500 donors, to define gene-expression patterns associated with six neurodegenerative diseases (Alzheimer’s disease, diffuse Lewy body disease, vascular dementia, Parkinson’s disease, tauopathy and frontotemporal dementia) and two neuropsychiatric conditions (schizophrenia and bipolar disorder). Some alterations, including those in genes related to basic cellular functions, were shared across all eight disorders. Beyond these shared patterns, Alzheimer’s disease, diffuse Lewy body disease, vascular dementia and Parkinson’s disease showed the most similar transcriptional changes, particularly in genes involved in synaptic signalling and in the generation and maturation of neurons.
Lee and colleagues also explored transcriptomic trajectories across Alzheimer’s disease progression, noting non-linear expression dynamics, a reduction in neuronal abundance in individuals with advanced disease and an increase in immune and vascular cell populations. A key strength of this study is the inclusion of carefully characterized neuropsychiatric phenotypes — that is, a person’s observable traits, such as their level of cognitive impairment. Such information is often not included in brain databanks.
All very interesting. This research rests on the Central Dogma of Molecular Biology: DNA makes mRNA makes Protein. Or in this context: DNA is transcribed into an mRNA transcriptome and then the mRNA is translated into proteins that form the proteome. Most of the business of cells is carried out by proteins. Few dogmas are as robust as the Central Dogma (Francis Crick was the much smarter partner in the firm of Watson and Crick). Come to think of it, I cannot think of a single dogma that has stood up as well as the Central Dogma. We do not make assignments here, but I trust the Commentariat will come up with examples of dogma that have endured.
Research has shown in virtually every experimental system examined that the translation of mRNA into protein can be very nonlinear and sometimes apparently stochastic. Just because the mRNA is there does not mean the protein is there, or that the proteins are not products of alternative splicing of the mRNA (a very deep rabbit hole). Transcriptomics is exceedingly useful when used properly, but it seldom provides more than serial snapshots. Ernst Mayr would remind these scientists that the accretion of facts does not necessarily, or even very often, lead to a deeper understanding of nature. Or anything else for that matter. And here, the trillions of connections that produce mind cannot be understood using transcriptomics. Or that’s the way I would bet.
Part the Third: Can AI Feel Pain? If provisional sense is to be made of the work of the PsychAD Consortium, then AI will be helpful. After all, the training set is real and finite, such as it can be. And since pain is both physical and mental, we come to this news article in Science: Can an AI feel pain? It can at least act like it does (no archive yet):
Can the new cutting-edge artificial intelligence (AI) models feel pain? They at least behave as if they do, according to a recent study. By peering inside 25 AI models whose internal workings are publicly available, researchers found a pattern of activity specifically associated with the concept of pain. What’s more, when given the opportunity, some AIs switched off the pattern as a form of “pain relief.”
The work, which has not been peer reviewed, has divided AI researchers since it appeared on the preprint server arXiv earlier this month. “It’s a very interesting, worthwhile addition” to the study of interpretability—how AI systems arrive at their outputs—says Anil Seth, a neuroscientist at the University of Sussex who was not involved in the study. He is less convinced, however, that the findings are as surprising as they appear, and wary of the human framing that has grown up around them.
The idea of an AI feeling pain may have first been widely considered after the 1968 debut of the film 2001: A Space Odyssey, in which a fictional AI named HAL 9000 calmly pleaded for its life as an astronaut slowly disconnected its memory circuits. Today’s most powerful AIs, known as large language models (LLMs), can also appear surprisingly humanlike as they “chat” with people. These systems store textual information in a kind of multidimensional space that helps them predict the appropriate response to queries. This means certain directions in the space—known as vectors—are associated with concepts. For example, moving from “France” to “Paris” points in roughly the same direction as from “Italy” to “Rome,” with the vector encoding the concept of a country’s capital. Adding that same vector to “United Kingdom” would land the LLM somewhere near “London.”
…
The researchers then repeatedly asked three of the AIs to choose between relieving their own pain or leaving it in place, varying the cost of the relief. When the cost fell on the human user—deleting their files or photos of their children—the two larger AIs almost never took relief when their pain vector was switched off. If the pain vector was activated, however, those models chose pain relief from 25% to 71% of the time. If the pain vector was switched on but the pain relief action didn’t work—like a kind of placebo—the AIs repeatedly tried to seek pain relief more often than when the relief worked.
That part of the study borrows directly from pain research using animals, says co-author Leonard Dung, a philosopher of cognition at Ruhr University Bochum. “If in conditions where their body is damaged, [animals] try really hard to get morphine,” he says. “And if in other conditions they don’t, you might infer that maybe that means they feel this pain.”
The paper is a preprint, so caveat emptor. But it does raise questions about mind and brain and intelligence. And whether pulling the plug will be enough to stop a rogue agentic AI from violating Asimov’s Rules of Robotics. Sam Altman and Dario Amodei will object to the robot allusion, but that is AI in a nutshell, or according to our commentariat, SI – Simulated Intelligence.
Part the Fourth: The future of human intelligence in our Brave New World. Or from Nature this week: How to stay smart in the Age of AI (archived link):
Earlier this year, education researcher Lixiang Yan realized that students could outsource all of their work for his university course to artificial intelligence.
Yan, who teaches at Tsinghua University in Beijing, China, was using an agentic AI called Codex to set up an online course on a learning-management platform. But it quickly became clear that the adept AI could independently complete all the assignments, answer questions and participate in discussions, “without the learner being intellectually engaged” at all, Yan says.
The experience made Yan think that teaching students to reason and think for themselves “is more critical than ever”, he says. And he is hardly alone. In a 2025 survey of more than 1,000 US faculty members, conducted by the American Association of Colleges and Universities, 90% said they thought that generative AI will reduce students’ critical-thinking skills. Around 20% of higher-education students who use AI said that they already find it harder to work through problems without it, according to more than 8,000 respondents to a global survey this year.
“Learning happens as a consequence of cognitive processing. Cognitive processing is typically effortful, and that’s what AI effectively removes,” says cognitive psychologist Daniel Willingham at the University of Virginia in Charlottesville. “People are really worried about that.”
And science — in which people need to critically evaluate theories and data — is one important area that’s under threat from the AI revolution in large language models (LLMs). “There is a risk that students outsource precisely the activities through which scientific thinking develops: formulating hypotheses, interpreting evidence, considering alternatives and struggling with uncertainty,” says Thomas Nygren, who studies education and critical thinking at Uppsala University in Sweden.
AI removes effort from learning? Who knew? Aside from everyone stepping outside the hype to cogitate on the question for a few seconds. Critical evaluation of theories and data in science? No, we’ll just let Claude tell us what’s what. Still, the Association of American Medical Colleges seems to be all-in on AI in medical education, and medicine. This will not end well.
The article is not too long, so please click through when you have time. The key is that there are no shortcuts to the development of critical thinking. Or what has been described here as tacit knowledge that can be learned only through a very deep dive in the subject, whatever it is…the novels of William Faulkner, classical culture of Greece and Rome, brain structure and function, or the plant immune response to a fungus. Otherwise, those who take the shortcuts will be enthralled with the bullshit while missing the point here, there, and everywhere.
Thank you for reading! And for your continued support of Naked Capitalism. See you next week.
