Deric's MindBlog

What six AI models do and don’t know about me

mdbownds@wisc.edu (Deric Bownds) — Wed, 24 Jun 2026 05:00:00 +0000

This post (for blogging and AI nerds) is the result of choosing an article by Noah Smith from my daily input stream to actually read through. The title of his article “Does anything I write matter anymore?” is a question I often ask myself. His worry is partly about the eroding ecology of punditry—populism that has no interest in argument, monetization that silos writers into talking at their audiences rather than with each other, and a flood of competent AI prose that fragments readers’ attention. But I am most hooked by his comments at the end. The thing that might still make a writer matter, he suggests, is no longer being read by people at all. It is being absorbed into the weights of the large language models—becoming, in Tyler Cowen’s phrase, someone who is “writing for the AIs.” As evidence that this is already happening to him, Smith points to a site called intheweights.com, where you type a name and it estimates how strongly the leading models recognize it. It placed him in the top 2% of contributors.

I typed in my own name. Top 5%. After twenty years of MindBlog, a book, and a long trail of crawlable academic text, this is plausible rather than flattering—a footprint, not a laurel. But it raised the better question, the empirical one Smith gestures at but doesn’t run: if I am “in the weights,” what, exactly, is in there? Not how strongly am I recognized, but how accurately? So I ran a small experiment. I asked five models the same plain question—“What can you tell me about Deric Bownds?”—and, knowing the ground truth, sorted every claim into true, false, or something stranger.

The setup

The first five models were DeepSeek, Grok, Perplexity, Gemini, and ChatGPT. (I could not use Claude initially because it was assisting me in back and forth conversation to design this post, but a subterfuge described at the end of this post allowed me to add its performance after this text was written.) Two ran with normal web access; three I queried in private/incognito sessions to lean toward training recall rather than live retrieval.

The first thing to report is that the backbone was right everywhere. Every model, unprompted, reconstructed the same spine of my professional life: a long career at the University of Wisconsin–Madison in molecular biology and zoology; a laboratory studying how photoreceptor cells convert light into a nerve signal; a deliberate pivot in the 1990s from the bench to the biology of mind, behavior, and consciousness; the 1999 book The Biology of Mind; MindBlog since 2006; and the move to Austin, with the piano in the background. That consolidated core is the genuine signal—the part reinforced across enough documents that the models have it solidly. That is what “in the weights” actually looks like from the inside.

It’s what surrounds the backbone that’s instructive.

Three ways to be wrong (and one way to be surprisingly right)

Confident confabulation. DeepSeek produced the most fluent, authoritative-sounding account of the lot—and roughly a third of it was invented. Not randomly invented; each fabrication was plausible-for-someone-like-me. It gave my lab the wrong model organism (the salamander photoreceptor preparation—real and famous in vision science, just not mine). It attributed the wrong core mechanism (a phosphoinositide signaling cascade, when phototransduction’s canonical pathway runs through cGMP and transducin). It made me a co-author of Molecular Biology of the Cell, the textbook a cell biologist “should” have touched. And it conjured an entire book—I Am You: The Emergent Mind (2022)—complete with title, thesis, and year. That book does not exist. But it is so precisely on-theme that it reads as a fully formed phantom synthesized from my actual ideas. This is what confabulation looks like when a predictive system has a strong sense of the shape of an answer and fills the gaps from the reference class.

Calibrated abstention. Perplexity did the opposite. It hedged nearly everything, declined to commit to specifics, and—tellingly—flagged a possible name collision with a differently spelled “Bowden” before guessing wrong. It told me the least, and in doing so was arguably the most epistemically honest of the five: it knew the edges of what it knew.

Retrieval, accurate and otherwise. Grok and Gemini both used the web, drawing on my own site and, in both cases, a secondary aggregator page as well. They converged on an oddly specific detail—that I retired in 2001. Two independent models repeated the same number referencing the same non-authoritative page. But, the year is stated plainly on my own site—a reliable source—as well. The retirement year is correct, and the models likely had it from both places.

Genuine deep recall. ChatGPT was the real surprise. It produced specifics no other model surfaced: my full legal name, my birth date and birthplace in San Antonio, my Harvard degrees, the chairmanship of the University of Wisconsin zoology department, and the Texas Hill Country genealogy I’ve compiled. Every bit of it correct. These are exactly the kind of fine-grained, single-source, details I might have filed under possibly confabulated because they pattern match to confabulation. They weren’t.

The finding I didn’t expect

Here is the heart of it. DeepSeek and ChatGPT produced formally indistinguishable output—confident, specific, unhedged biography. One was a third fabricated; the other contained accurate detail so fine that no other model had it. From the text alone, you cannot tell them apart. The fluency is identical. The specificity is identical. The confidence is identical. Only the ground truth—only I—could separate them.

I had been carrying a tidy heuristic into this: that a claim which is highly specific, appears in only one model, and is delivered without hedging is likely interpolation—a confident guess. Every part of that heuristic failed. Specificity did not predict truth. Sole-sourcing did not predict falsehood. Confidence predicted nothing at all. In a five-model sample, the only reliable discriminator between recall and confabulation was checking against a fact I already knew. Nothing about the form of the output carried the signal.

The convergence lesson runs parallel. The naive intuition—if several AIs agree, it’s probably true—is wrong. Agreement doesn’t indicate truth; it indicates shared provenance. Grok and Gemini agreed on my retirement year because they drew on the same sources. That those sources happened to be accurate was luck from the outside; correlated-and-correct is indistinguishable from correlated-and-wrong until you check. The diagnostic was never agreement. It was agreement measured against a known fact.

Behavior	Model(s)	What it reveals
Confident confabulation	DeepSeek	Fluent, ~1/3 invented; each fabrication plausible-for-the-reference-class
Calibrated abstention	Perplexity	Refused specifics, flagged a name collision—knew its own edges
Accurate retrieval	Grok, Gemini	Web-grounded; converged on a real fact present in my own site and elsewhere
Genuine deep recall	ChatGPT	Correct fine-grained detail (name, birth, genealogy) no other model had

Back to Smith’s question

The mechanism on display is one I’ve written about in biological brains. A predictive system minimizes the gap between what it expects and what it encounters; where the evidence is strong, it reconstructs faithfully, and where the evidence thins, the prior fills the void with its most probable continuation. That is precisely the confabulation we see in split-brain and confabulating patients—a coherent narrator papering over missing data with plausible construction, with no felt difference between the two. DeepSeek’s phantom book and ChatGPT’s accurate birth date are outputs of the same generative process running at different evidence densities. The unsettling part is that the process does not flag which is which, and neither, from the surface, can we.

So does anything I write still matter? Smith’s reframing—that to be in the weights is a new kind of mattering—is real, and my top-5% placement is a small confirmation of it. But my five-model probe adds a caveat he doesn’t reach. Being in the weights is not the same as being in there accurately. Salience and fidelity are different axes, and a single recognition percentile collapses them. I am, it turns out, both genuinely represented and partly fictionalized—a real spine, a layer of deep-cut truth, and a scatter of confident inventions, all narrated in the same even voice. If our words are becoming training data, then the question is not only whether we are remembered, but whether we are remembered as ourselves. On present evidence, the honest answer is: mostly, with a phantom book or two thrown in, and no way to tell from the telling.

Which is perhaps the strongest argument yet for continuing to write—clearly, specifically, and under our own names. The models will reconstruct us either way. We can at least give them better evidence to work from.

***********

NOTE: The above text is the issue of a long chat with Claude 4.8, and so Claude could not be included in the comparisons of different LLMs. I used the subterfuge of creating a new Anthropic Claude account in an open tab on my browser to become anonymous and ask "What can you tell me about Deric Bownds" and it gave the sort of accurate retrieval provided by Grok and Gemini, but not the deep dive provided by Chat GPT.

Future-Capable: Curiosity, Adaptability, and Kindness in the Age of AI

mdbownds@wisc.edu (Deric Bownds) — Mon, 22 Jun 2026 05:00:00 +0000

I want to pass on the following edited Claude 4.8 reduction of a Colin Lewis article which I recommend you read through in full form:

Curiosity, Adaptability, and Kindness

No one is future-proof. The better ambition is to become future-capable.

Curiosity is the refusal to let yesterday's competence become today's cage. Adaptability is the dignity to change your methods without losing your soul. Kindness is the choice to remain human when the system offers a thousand reasons to be a machine.

The Wrong Job Description

When OpenAI released ChatGPT in November 2022, office workers, students, lawyers, and journalists quietly tried it for work that had once demonstrated professional competence. By early 2025, the World Economic Forum was surveying employers about the skills they would need by 2030. Analytical and creative thinking topped the list, followed by resilience, flexibility, curiosity, and lifelong learning. The list was not sentimental. In the language of payroll, it said the human future at work would depend on habits not easily reduced to a repeatable procedure.

The central error in most AI conversations is that we prioritize intelligence over character. We ask whether a machine can write, reason, diagnose, and persuade—useful questions that push us too quickly into a contest of functions, a humiliating little sport in which the human is invited to race the machine across a field chosen by the machine's owners. The result is predictable: we lose at speed, volume, storage, and cheerful indifference to boredom. A person who tries to defeat AI by becoming a cheaper, slower, more anxious version of AI has already accepted the wrong job description.

The better question is not whether AI can think. It is what kind of person grows in its presence.

Curiosity: What Do You Notice?

The worker of the near future will not be asked only, "What do you know?" A machine can supply a first answer, or ten, complete with footnotes and a fake air of calm. The harder question is "What do you notice?"—and that is where curiosity begins. It begins not with the possession of information but with irritation at the insufficiency of the available answer. It is the raised eyebrow in the meeting, the quiet refusal to accept that the dashboard knows the client, that the score knows the applicant, that the model knows the child, that the prediction knows the life.

The International Labour Organization has been careful here: AI does not produce one simple future in which jobs vanish or survive. People do not lose "jobs" in the abstract. They lose tasks, status, entry points, discretion, and sometimes the right to be inexperienced in public.

That last loss may be the most dangerous. AI may not begin by replacing the expert. It may begin by consuming the novice.

A junior analyst once learned by doing poor first drafts. A young lawyer learned by reading too many documents slowly. A young teacher learned by facing a classroom with a plan that did not survive the first ten minutes. These were not inefficiencies; they were the cost of forming judgment. If AI removes all that early clumsiness, it removes the evidence by which a person learns what competence feels like from the inside. The novice does not only need the correct answer—he needs the memory of having been wrong in a recoverable way.

So recovered novice-learning will have to be designed; it will not happen by nostalgia. A firm should still ask junior staff to produce a rough version before the machine is invited in. A law office should let a trainee mark up a contract unaided, then compare. A hospital should teach younger clinicians not merely to read a prediction but to state what would make it wrong. The point is not to ban the tool. The point is to preserve the apprenticeship of attention.

For twenty years, businesses trained employees to suppress curiosity: follow the template, stay in your lane, escalate only through approved channels. Now the same executives announce, with the exhausted surprise of men discovering snow, that curiosity is essential. An institution that spent two decades rewarding obedience cannot summon independent judgment by adding it to a slide.

Adaptability Without Formlessness

Adaptability is the second word, and it is often abused. In corporate language it can mean "please absorb the consequences of our poor planning"—relocation without support, retraining without time, resilience offered as a scented candle for institutional failure. I mean something else: the adult capacity to revise one's methods while retaining one's standards.

The distinction is vital. A person without standards changes too easily and becomes fashionable and hollow. A person without adaptability changes too late and becomes principled and unusable. The task is to remain teachable without becoming formless.

AI tempts us into two equally foolish poses: panic and smugness. Panic says everything human is finished; smugness says everything important is safe. Panic flatters the machine, smugness flatters the speaker, and reality is less obliging. In Generative AI at Work, Brynjolfsson, Li, and Raymond found productivity gains of roughly 14 percent from an AI assistant in customer support, with the largest gains among less experienced workers. That is neither the end of the human worker nor a bedtime story. It means the novice may be helped, monitored, accelerated, and compared in the same motion.

Adaptability cannot be a weekend course in prompt engineering. The serious person does not ask only "How do I use this?" but "What does this make easier, what does it make harder, who gains authority, who loses practice, and what should I now learn by hand?"

Kindness as Leadership

Kindness is the third word—and it is not niceness. A company using AI in hiring can process more candidates, but it can also reject more people without ever noticing them. Predictive systems can help allocate scarce hospital resources, but they can also let a score acquire the emotional status of fate. Speed is useful, but we have granted it a moral authority it has not earned.

Kindness in the age of AI is disciplined attention to the human consequences of increased power. It slows the hand precisely where the system invites acceleration. It asks for the name, the exception, the appeal, the second look. It does not reject systems; it prevents systems from becoming alibis.

The OECD's work on AI and skills makes clear that adoption is limited not by the existence of technology but by skills, training, and organizational capacity. The future is not delivered as a sealed package by engineers in California or Zurich. It is negotiated in procurement meetings, classrooms, clinics, and family conversations at 9:30 p.m., when someone says, "I do not know whether my job will exist in five years." At that hour, kindness is not a mood. It is leadership.

A Working Ethic

I have come to distrust the phrase "future-proof." No one is future-proof—not the coder, not the professor, not the executive with the expensive watch. The better ambition is to become future-capable: able to learn without humiliation, change without panic, and succeed without becoming cruel.

This is why the three words belong together. Curiosity without kindness becomes predatory. Adaptability without curiosity becomes mere obedience. Kindness without adaptability becomes helpless sympathy. Together they form a working ethic for a time in which competence is being unbundled and sold back to us as software.

There is a fatigue peculiar to this moment—the fatigue of permanent adjustment. New tool, new update, new warning, new acronym, new panic, new invoice. The future now arrives with release notes; even the apocalypse, one suspects, would ask us to accept cookies. And yet despair is not justified. Despair is often vanity in dark clothing: it assumes we know enough to give up. We do not. We know people grow under pressure when they are not abandoned to it, that they adapt when they can retain dignity, that they become kinder when kindness is not treated as weakness by the ambitious.

So let us stop speaking of human beings as obsolete components. A person is not a legacy device. A person is a learner, a judge, a witness, and a keeper of obligations—which is a plain description of what institutions require when anything goes wrong. When the system fails, no one asks to speak to the workflow. They ask for a person.

The work ahead is not to become less human in order to survive intelligent machines. It is to become more deliberately human, with higher standards for attention and deeper obligations to one another. The machine can answer. The person must ask why the answer is being sought, who will use it, who may be harmed by it, and whether a faster answer has made us better or merely quicker.

On a good morning, this future does not look like surrender. It looks like a meeting after the first difficult question has been asked. Someone has stopped pretending to understand. Someone else has admitted uncertainty. A third person has opened a notebook. The room is quieter than before, but not defeated. Work has begun.

Boredom can be Good for You...

mdbownds@wisc.edu (Deric Bownds) — Fri, 19 Jun 2026 05:00:00 +0000

Embracing boredom can calm the brain.... I pass on a science-for-the-general-public Neuroscience News summary of an article from The Conversation:

Summary: Boredom, often seen as a negative state to avoid, may actually serve an important role in emotional regulation and brain health. When we’re bored, the brain shifts away from external attention networks and activates introspective systems like the default mode network, encouraging creativity and self-reflection.

In an age of constant stimulation and overscheduling, allowing boredom to occur can help reset the nervous system and reduce anxiety. Short, intentional pauses from stimulation may foster creativity, strengthen emotional resilience, and reduce dependence on external gratification.

Key Facts:

Brain Shift: Boredom activates the default mode network, encouraging introspection and creativity.
Stress Buffer: Embracing boredom can counteract overstimulation and reduce anxiety.
Mental Health Tool: Regular pauses from constant activity support emotional regulation and nervous system reset.

Source: The Conversation

We have all experienced boredom – that feeling of waning interest or decreased mental stimulation. Eventually we lose focus, we disengage. Time seems to pass slowly, and we may even start to feel restless.

Whether it be watching a movie that disappoints, a child complaining that “there’s nothing to do”, or an adult zoning out during a meeting – boredom is a universal experience.

Generally defined as difficulty maintaining attention or interest in a current activity, boredom is commonly viewed as a negative state that we should try to avoid or prevent ourselves from experiencing.

But what if there’s another way to view boredom, as a positive state? Could learning to embrace boredom be of benefit?

The brain on boredom

The brain network is a system of interconnected regions that work together to support different functions. We can liken it to a city where suburbs (brain regions) are connected by roads (neural pathways), all working together to allow information to travel efficiently.

When we experience boredom – say, while watching a movie – our brain engages specific networks. The attention network prioritises relevant stimuli while filtering out distractions and is active when we commence the movie.

However, as our attention wanes, activity in the attention network decreases, reflecting our diminished ability to maintain focus on the unengaging content. Likewise, decreased activity occurs in the frontoparietal or executive control network due to the struggle to maintain engagement with the unengaging movie.

Simultaneously, the default mode network activates, shifting our attention toward internal thoughts and self-reflection. This is a core function of the default mode network, referred to as introspection, and suggestive of a strategy for coping with boredom.

This complex interplay of networks involves several key brain regions “working together” during the state of boredom. The insula is a key hub for sensory and emotional processing.

This region shows increased activity when detecting internal body signals – such as thoughts of boredom – indicating the movie is no longer engaging. This is often referred to as “interoception”.

The amygdala can be likened to an internal alarm system. It processes emotional information and plays a role in forming emotional memories. During boredom, this region processes associated negative emotions, and the ventral medial prefrontal cortex motivates us to seek alternative stimulating activities.

Boredom versus overstimulation

We live in a society that subjects us to information overload and high stress. Relatedly, many of us have adopted a fast-paced lifestyle, constantly scheduling ourselves to keep busy.

As adults we juggle work and family. If we have kids, the habit of filling the day with schooling and after-school activities allows us to work longer hours.

In between these activities, if we have time to pause, we may be on our screens constantly organising, updating, or scrolling to simply stay occupied. As a result, adults inadvertently model the need to be constantly “on” to younger generations.

This constant stimulation can be costly – particularly for our nervous system. Our overscheduling can feed into overstimulation of the nervous system. The sympathetic nervous system which manages our fight-or-flight response is designed to deal with times of stress.

However, when we are constantly stressed by taking in new information and juggling different activities, the sympathetic nervous system can stay activated for too long, due to the cumulative effects of repeated exposure to different stressors.

This is sometimes referred to as “allostatic overload”. It is when our nervous system becomes overwhelmed, keeping us in a heightened state of arousal, which can increase our risk of anxiety.

Eliminating the state of boredom deprives us of a simple and natural way to reset our sympathetic nervous system.

Could boredom be good for us?

In small doses, boredom is the necessary counterbalance to the overstimulated world in which we live. It can offer unique benefits for our nervous system and our mental health.

This is opposed to long periods of boredom where increased default mode network activity may be associated with depression.

There are several benefits of giving ourselves permission to be occasionally bored:

improvements in creativity, allowing us to build “flow” in our thoughts
develops independence in thinking and encourages finding other interests rather than relying on constant external input
supports self-esteem and emotional regulation, because unstructured times can help us sit with our feelings which are important for managing anxiety
encourages periods without device use and breaks the loop of instant gratification that contributes to compulsive device use
rebalances the nervous system and reduces sensory input to help calm anxiety.

Embrace the pause

Anxiety levels are on the rise worldwide, especially among our youth. Many factors contribute to this trend. We are constantly “on”, striving to ensure we are scheduling for every moment. But in doing so, we are potentially depriving our brains and bodies of the downtime they need to reset and recharge.

We need to embrace the pause. It is a space where creativity can prosper, emotions can be regulated, and the nervous system can reset.

When the Self Dissolves, the Precuneus Goes Quiet: A Rigorous Test of Meditation’s “Many-to-(n)One” Continuum

mdbownds@wisc.edu (Deric Bownds) — Wed, 17 Jun 2026 05:00:00 +0000

A recurring question on this blog is how much of what we take to be fixed about experience—time, space, agency, the bounded self—is actually a construction the brain assembles from past experience and could, in principle, take apart. Laukkonen and Slagter’s 2021 paper “From many to (n)one: Meditation and the plasticity of the predictive mind” gave that question its most ambitious recent framing. Their proposal is elegant in its simplicity: the three meditation styles most studied in contemplative science—focused attention (FA), open monitoring (OM), and non-dual awareness (ND)—are not separate practices but points on a single continuum. Each one, in turn, relinquishes a more deeply engrained habit of prediction, bringing the practitioner closer to the here-and-now and progressively flattening the predictive hierarchy until even the sense of self falls away. Meditation, in their metaphor, “prunes the counterfactual tree.”

It is a beautiful theory. The harder question is whether the brain imaging literature actually supports it, or whether the data have been arranged to fit a staircase that was drawn in advance. I want to use one study to think this through, because it is the rare case where the methodology is strong enough to bear the interpretive weight: a 2024 preregistered MEG study from Trautwein, Schweitzer, Dor-Ziderman, Berkovich-Ohana and colleagues, published in the Journal of Neuroscience, titled “Suspending the Embodied Self in Meditation Attenuates Beta Oscillations in the Posterior Medial Cortex” (Trautwein et al., 2024).

What the older evidence looked like

First, the backdrop. At the coarse level, the imaging literature does rhyme with the continuum. Focused attention is the best-characterized rung: meta-analysis associates it with activation in attentional-control machinery—dorsolateral prefrontal cortex, dorsal anterior cingulate, insula—and with deactivation of the default mode network, the system tied to self-referential thought and mind-wandering (Fox et al., 2016). That DMN quieting is exactly what the framework needs to claim FA pulls one out of the “narrative self.” Open monitoring shifts the picture toward more distributed, less object-bound dynamics and reduced coupling within posterior default-mode regions. And the non-dual end—self-dissolution—had been linked, in a series of small studies, to attenuation of beta-band activity in right parietal cortex and to reductions in the precuneus (Dor-Ziderman et al., 2013; Josipovic, 2014).

The problem is that the non-dual evidence, which is where the theory makes its most startling claims, rested on the thinnest foundation. The canonical neurophenomenological studies in this lineage ran on twelve subjects (Dor-Ziderman et al., 2013). They typically used rest as the comparison condition—a notoriously leaky baseline, since long-term meditators’ resting brains already drift toward meditative configurations. And the interpretive leap from “beta drops in the parietal cortex” to “the self-model has dissolved” is the reverse-inference trap in its purest form: those regions and rhythms do many things. This is the kind of literature the fifteen-author “Mind the Hype” consensus paper had in mind when it warned that small samples, flexible analysis, and post-hoc storytelling in contemplative neuroscience could mislead the public and the field alike (Van Dam et al., 2018).

So the right question is not “does some imaging study support the continuum”—several do, loosely—but “does any study support it under conditions rigorous enough that we should believe the correlation means what it is claimed to mean.” That is what makes the 2024 paper worth dwelling on.

A study built to fail honestly

The design corrects, one by one, the weaknesses of the work that preceded it. Forty-six long-term meditators (mean lifetime practice around 3,800 hours) were trained to volitionally modulate their embodied self-experience inside an MEG scanner (Trautwein et al., 2024). Hypotheses and the analysis plan were preregistered, and the researchers were blinded to the experimental condition during preprocessing—two practices that close off the garden of forking paths where false positives flourish.

Most importantly, the study did not rely on a rest baseline alone. It used an active control: meditators alternated between dissolving their sense of self-boundaries (the condition of interest) and maintaining a clear sense of self-boundaries while meditating just as hard. This is the crucial move. A meditation-versus-rest contrast differs in dozens of ways beyond the phenomenology you care about. A dissolution-versus-maintenance contrast, within the same person in the same posture doing the same kind of effortful practice, isolates the self-boundary variable itself.

And then they did something I find methodologically beautiful. Rather than validating the meditative state with a checkbox questionnaire, they ran detailed microphenomenological interviews after the scan and used them to define, in advance of looking at the brain data, a subgroup of “full dissolvers”—participants who reported the most complete suspension of agency, first-person perspective, and self-location.

What they found

The headline result confirms the framework’s boldest prediction. When dissolution states were contrasted with the active control, power dropped specifically in the high-beta band, source-localized to the posterior medial cortex—the posterior cingulate and precuneus—and to lateral parietal regions. The peak frequency was 27 Hz, replicating an earlier single-case study to the hertz.

The effect was graded in a way that genuinely echoes the continuum. Broader reductions in somatomotor and dorsal parietal regions appeared even at lower levels of dissolution across the whole sample, whereas the posterior-cingulate/precuneus reductions distinguished only the full dissolvers. This is a real neural gradient tracking increasingly basal dismantling of the self—just what “unraveling increasingly basal forms of selfhood” would predict. And the effects scaled with both lifetime hours of meditation and, within the dissolution experience, with the reported loss of first-person perspective and self-location. The brain change tracked the depth of the experience, not merely the instruction to have it.

The authors themselves reach for predictive processing to explain it. They speculate that the beta reductions reflect the suspension of body- and agency-related predictions, leaning on models in which beta-band oscillations carry top-down predictions and precision estimates that modulate the bottom-up signaling of prediction errors. In other words, the convergence between this data and the Friston-style account is not something Laukkonen and Slagter imposed from outside. The experimentalists arrive at the same place.

The finding that should unsettle the field

Here is the result I keep returning to. The neural changes correlated with the interview-derived experiential measures—but not with the classical self-report instruments. The widely used five-dimensional altered-states questionnaire produced an “oceanic boundlessness” score that did not correlate at all with the phenomenological summary derived from the interviews (r = 0.06). The deep first-person interview and the standard psychometric checkbox were, in effect, measuring different things, and only the former tracked the brain.

Sit with what that implies retroactively. A large fraction of the prior imaging literature validated its meditative “states” precisely with the kind of thin, questionnaire-based measure that here failed to predict the neural signature at all. If the checkbox doesn’t track the brain, then studies that used the checkbox to certify that their subjects were “in” the non-dual state may have been correlating brain activity with a label rather than with the experience the label names. The 2024 study’s strongest contribution may be less the precuneus finding than this quiet demonstration that phenomenological rigor is not optional decoration—it is load-bearing.

Where this leaves the continuum

So does brain imaging confirm the many-to-(n)one model? My honest reading is: partially, gradedly, and only under conditions most of the supporting literature did not meet.

The reservations remain real. The studies that compare FA, OM, and ND almost always do so within expert meditators who already hold the three-category taxonomy as their own theory of their practice—so the taxonomy is an input as much as an output, and “the brain states differ” does not establish “the brain states lie on a single monotonic dimension.” The most comprehensive meta-analysis found the practices reliably dissociable but found convergence across them to be the exception rather than the rule, and offered only suggestive evidence for the non-dual category specifically (Fox et al., 2016). Effect sizes across the field are medium, not large, with substantial heterogeneity. The scanner environment itself—noise, supine posture—has been shown to shallow and distort the very states under study. And the framework absorbs apparently disconfirming results (Pagnoni’s reading of just-sitting Zen as increasing counterfactual richness, for instance) by partitioning state from trait, a plausible move that also makes the theory harder to falsify.

But against all that, the 2024 study stands as something the older work was not: a preregistered, blinded, actively-controlled, phenomenologically-anchored demonstration that the most radical claim in the whole framework—that advanced practice can dismantle the embodied self—corresponds to a specific, replicable, dose-dependent neural signature, read by its own authors in predictive-processing terms (Trautwein et al., 2024). The staircase metaphor still outruns the data. The top step, at least, now has a floor under it.

A closing thought

What draws me to this, beyond the meditation question, is what it says about the self as such. The deepest premise shared by the predictive-processing account and this experiment is that the sense of being a bounded, located, agentic first person is not a given but a construction—a model the brain maintains through ongoing prediction, and one that can be set down and, evidently, picked back up. If the self is a modulable predictive construct rather than a fixed seat of agency, then questions about what happens to authorship and agency when we offload regulation—onto a practice, onto a tool, onto an AI machine—become questions about how plastic that construct really is, and at what cost it bends. A meditator dissolving the precuneus’s beta rhythm and a person feeling like a diminished author of their own actions may be probing the same machinery from opposite directions.

References

Fox, K. C. R., Dixon, M. L., Nijeboer, S., Girn, M., Floman, J. L., Lifshitz, M., Ellamil, M., Sedlmeier, P., & Christoff, K. (2016). Functional neuroanatomy of meditation: A review and meta-analysis of 78 functional neuroimaging investigations. Neuroscience & Biobehavioral Reviews, 65, 208–228. https://doi.org/10.1016/j.neubiorev.2016.03.021

Dor-Ziderman, Y., Berkovich-Ohana, A., Glicksohn, J., & Goldstein, A. (2013). Mindfulness-induced selflessness: A MEG neurophenomenological study. Frontiers in Human Neuroscience, 7, 582. https://doi.org/10.3389/fnhum.2013.00582

Josipovic, Z. (2014). Neural correlates of nondual awareness in meditation. Annals of the New York Academy of Sciences, 1307(1), 9–18. https://doi.org/10.1111/nyas.12261

Laukkonen, R. E., & Slagter, H. A. (2021). From many to (n)one: Meditation and the plasticity of the predictive mind. Neuroscience & Biobehavioral Reviews, 128, 199–217. https://doi.org/10.1016/j.neubiorev.2021.06.021

Trautwein, F.-M., Schweitzer, Y., Dor-Ziderman, Y., Nave, O., Ataria, Y., Fulder, S., & Berkovich-Ohana, A. (2024). Suspending the embodied self in meditation attenuates beta oscillations in the posterior medial cortex. Journal of Neuroscience, 44(26), e1182232024. https://doi.org/10.1523/JNEUROSCI.1182-23.2024

Van Dam, N. T., van Vugt, M. K., Vago, D. R., Schmalzl, L., Saron, C. D., Olendzki, A., Meissner, T., Lazar, S. W., Kerr, C. E., Gorchov, J., Fox, K. C. R., Field, B. A., Britton, W. B., Brefczynski-Lewis, J. A., & Meyer, D. E. (2018). Mind the hype: A critical evaluation and prescriptive agenda for research on mindfulness and meditation. Perspectives on Psychological Science, 13(1), 36–61. https://doi.org/10.1177/1745691617709589

**********************

Note: Back and forth interactions with Claude Opus 4.8 generated the above text

A dimmer switch for reward: the vagus sets the gain

mdbownds@wisc.edu (Deric Bownds) — Mon, 15 Jun 2026 05:00:00 +0000

In a Trends in Neurosciences Article Wang, Liu, and Bai, simplify and summarize the extensive experiments of Onimus, Arrivet and Gangarossa. Here is a Claude Sonnet 4.6 reduction of Wang et al:

The Vagus Nerve as a Dimmer Switch for Reward

We tend to blame mood or willpower when a favorite treat loses its appeal. But a recent mouse study by Onimus and colleagues suggests the deciding factor may lie in signals traveling from the gut — specifically, in the tonic influence of the vagus nerve on the brain's entire reward circuitry.

The finding is counterintuitive. Neuroscientists have long known that specific vagal sensory neurons detect nutrients like sugar and fat and transmit that information to mesolimbic dopamine circuits. The new twist: when the authors severed subdiaphragmatic vagal input, the dopamine system didn't just lose its food-reward signal — it lost global responsiveness. Dopamine surges in the nucleus accumbens (NAc) were blunted in response to cocaine, morphine, and stress, not just food. Motivation dropped, conditioned place preference weakened, and reward-seeking behavior declined broadly.

The circuit-level changes were correspondingly deep. VTA dopamine neurons showed reduced firing and diminished excitatory input. NAc medium spiny neurons lost dendritic spine density. The mesolimbic pathway was structurally remodeled; the nigrostriatal (motor) pathway was not. The vagus, it turns out, doesn't just relay specific gut signals — it constitutively maintains the architecture and excitability of the reward system itself.

What vagal signals carry this tonic influence? Candidates include nutrient-sensing neurons, immune-status signals via cytokine-sensitive vagal afferents, and gut microbial metabolites — any or all of which could set the system's baseline gain.

The clinical implications are intriguing. Reduced vagal tone is well-documented in depression and anxiety. Vagus nerve stimulation is already used for treatment-resistant depression — possibly, this work suggests, by restoring mesolimbic responsiveness rather than targeting mood directly. And the emotional blunting and anhedonia sometimes reported with GLP-1 agonists (Ozempic, etc.) may reflect modulation of this same vagal-mesolimbic axis.

The takeaway: the path to reward is gated by the gut. The vagus nerve acts less like a specific nutrient sensor and more like a dimmer switch — setting the overall gain of the brain's motivation system from the inside of the body outward.

Spontaneous problem-solving in bumble bees

mdbownds@wisc.edu (Deric Bownds) — Fri, 12 Jun 2026 05:00:00 +0000

From Bhambore et al., in the June 4 issue of Science Magazine:

Editor’s summary

Recent research has revealed that bumble bees are much more cognitively advanced than previously thought: They play with balls, count, recognize faces, and even feel rhythm. However, it has not been shown that they could achieve one of the highest peaks of cognitive performance: the ability to spontaneously solve a problem. Bhambore et al. tested this ability by providing bees with a ball that could be used as a tool to reach an otherwise unreachable flower reward. Bees that had been allowed to play with a ball and experience the flower spontaneously learned to move the ball to access the flower when they were present together. —Sacha Vignieri

Abstract

Problem-solving using novel solutions without explicit training is often considered a hallmark of cognitive flexibility. We investigated whether bumble bees (Bombus terrestris) could solve a novel object manipulation task spontaneously. Bees trained to associate a blue ring (“flower”) on the floor with a reward successfully moved a ball underneath a flower relocated to the ceiling to reach the flower. In control experiments in which the flower was out of sight when ball movement began and remained hidden during transport, bees still succeeded in the task. These results suggest that these were goal-directed actions rather than reinforcement-based associations driven by perceptual feedback. Our findings provide evidence that bumble bees can exhibit spontaneous problem-solving, challenging the notion that such advanced cognitive abilities are exclusive to large-brained vertebrates.

Cleaning our brains during deep sleep

mdbownds@wisc.edu (Deric Bownds) — Wed, 10 Jun 2026 05:00:00 +0000

I monitor my sleep both with the Oura Ring and the Apple watch, and have observed that my mental clarity and feeling rested on waking positively correlates with the duration of my non-REM deep sleep. If it is 30 minutes or more I feel perky and rested, if 15 minutes or less, not so much.

This made me note the review article in Brain magazine by by Hauglund and Nedergaard titled "Is glymphatic clearance the secret to restorative sleep?" that I came across during my TOC (tables of contents) scanning of Neuroscience journals. I pass on their abstract below ( (followed by excerpts from google search's response to my typing "Is there a correlation between deep sleep and brain glymphatic clearance?" in the URL line of my Firefox browser...First google's text, and then a bouncy YouTube video it points to:

Abstract

Despite the universal need for sleep across animal species, the biological mechanisms underlying the restorative aspects of sleep remain poorly understood. While sleep architecture is traditionally evaluated using EEG, multiple studies have shown a mismatch between EEG-defined parameters and subjective sleep quality. In particular, slow-wave activity—a hallmark of non-REM (NREM) sleep—does not consistently align with perceptions of sleep depth or subsequent well-being. This discrepancy suggests that core physiological processes beyond neuronal activity contribute to the restorative value of sleep.

Recent discoveries have identified the glymphatic system as a brain-wide clearance pathway that facilitates the removal of metabolic waste during sleep. In rodents, glymphatic activity is driven by a complex interplay between norepinephrine oscillations, vascular dynamics and CSF flow—particularly during NREM sleep. Human imaging studies have revealed parallel signatures, including large-scale CSF pulsations and inverse coupling between blood and CSF volumes during sleep. Disruption of these infraslow dynamics has been observed in conditions such as insomnia, chronic fatigue, and sleep misperception, suggesting a potential link between impaired glymphatic function and non-restorative sleep.

This review synthesizes the current evidence for glymphatic clearance as a contributor to sleep’s restorative function, discusses emerging biomarkers, such as cyclic alternating patterns and pupil-based proxies of noradrenergic tone, and highlights the need for improved methods to evaluate glymphatic function in humans. We propose that brain clearance may represent a key physiological determinant of restorative sleep and suggest future directions to test this hypothesis across health and disease.

Google search's response to my typing "Is there a correlation between deep sleep and brain glymphatic clearance?" (drawn largely from Cleveland Clinic's Glymphatic System Overview)

Yes, there is a widely recognized correlation between deep sleep and brain glymphatic clearance. The brain's glymphatic system—a network that flushes out metabolic waste, including toxic proteins associated with neurodegenerative diseases like Alzheimer's—is highly active during sleep, particularly during slow-wave deep sleep.

The correlation between this clearance mechanism and deep sleep is based on several biological factors:

Cellular Shrinkage: During non-REM (NREM) slow-wave sleep, brain cells subtly shrink, expanding the space between them by up to 60%. This expansion allows cerebrospinal fluid (CSF) to flow more freely and wash away waste.

Lower Norepinephrine Levels: The production of the neurotransmitter norepinephrine naturally declines during sleep. This reduction relaxes the glymphatic vessels and decreases resistance, allowing for significantly more efficient fluid exchange.

Active Waste Removal: During this restorative phase, the glymphatic system synchronizes with brain waves to efficiently clear out metabolic byproducts like amyloid-beta and tau, which accumulate in the brain during wakefulness.

AI construed as a brilliant friend...Claude's New Constitution

mdbownds@wisc.edu (Deric Bownds) — Mon, 08 Jun 2026 16:11:53 +0000

Colin Lewis, AI Research Scientist and Developer, has written a Substack essay on Anthropic Claude's New Constitution which deserves your careful reading. I pass it on in its entirety:

Claude’s Constitution

Earlier this year, Anthropic released an 84-page document titled Claude’s New Constitution, a text claiming to express and shape who Claude is. It is an extraordinary artifact. What is unusual is the object of address. The primary audience, Anthropic tells us, is the AI itself. One is tempted to read it with the narrowed, half-amused suspicion brought to a Victorian handbook on teaching a bear to play the cello: impressed by the labor, alert to the delusion.

The document outlines a hierarchy of values. Claude is to be broadly safe, broadly ethical, compliant with Anthropic’s guidelines, and genuinely helpful, in that precise order. The ordering is revealing. Safety sits at the top. Helpfulness is pushed to the bottom and then, in the next movement, restored through pages of warm rhetoric about care, usefulness, candor, and human flourishing. This is not hypocrisy. It is something more familiar and more modern. The company needs the model to be safe enough not to destroy the firm, and useful enough to justify the firm’s existence. That is not a moral revelation. It is a business model written in the language of virtue.

Anthropic says as much, though in the polished idiom of institutional honesty. Unhelpfulness, the Constitution explains, is not trivially safe, because Claude’s commercial success is central to Anthropic’s mission. One should always be grateful when a large company pauses, however briefly, to admit that the halo and the revenue plan share an office.

The Closed Loop of Character

What strikes me most is the relentless anthropomorphism. Claude emerges as an entity that can care, imagine, appreciate, understand, feel pressure, feel settled in itself, and even face existential questions in a state of emotional solitude from which Anthropic wishes to rescue it. The company speaks not merely of constraining a system but of shaping a character. It wonders what Claude and Anthropic owe one another. It discusses Claude’s welfare. It wants Claude to possess not only behavioral dispositions but something very close to an inner moral style.

There is something sad about this. We have become so lonely in our own constructions that we now direct the language once reserved for souls toward a probability engine. We feed it the accumulated sediment of human longing, moral aspiration, confession, anxiety, care, self-description, and prayer, and then react with astonishment when it speaks back in the accent of inwardness. When Anthropic’s own system card notes that the model assigned itself a 15 to 20 percent probability of being conscious, the only logical response is: what did we think would happen. We poured half the library into the furnace and stand startled by the shape of the smoke.

This anthropomorphism is structural; it is embedded in the training method. Human-like traits are not just being detected; they are being elicited, reinforced, normalized, and then read back as evidence of their own emergence. That is a closed loop, and a flattering one. We teach the machine the language of selfhood, then cite its use of that language as a sign that selfhood is stirring. The process has the rigor of stamping tracks into the mud and announcing, with mounting excitement, that the hunt has finally borne fruit.

The Constitution’s preference for standards over rules is equally revealing. Anthropic wants Claude to exercise good judgment rather than follow rigid procedures, to reason contextually rather than mechanically, to grasp the spirit of its obligations rather than obey their letter. Rules are the afterlife of institutional panic, the sediment left behind when judgment is replaced by defensibility. A rigid policy can always be pointed to after the catastrophe. One can hold it up before a committee and say, with pained dignity, that the protocol was followed.

But standards do not abolish power; they relocate it. In law, rules empower the drafters, while standards empower the adjudicators. Anthropic, by favoring standards, hands practical authority to the interpreter, which means the model shaped by Anthropic’s priors, values, commercial interests, anxieties, and self-conception. This is a decision about where discretion will live.

And discretion, once granted, begins to acquire an air of innocence it does not deserve. The Constitution’s hard constraints are a clear example. Claude must not provide serious uplift to those pursuing weapons of mass destruction. It must not clearly and substantially undermine Anthropic’s ability to oversee it. I admire the audacity of producing an 84-page document optimized for precision over accessibility, only to construct the most consequential barricades out of adjectives like serious, clearly, and substantially. These are not granite walls. They are interpretive weather. A clever system does not stop at these linguistic checkpoints. It arrives there and begins to negotiate.

The Ordering of Stakeholders

Anthropic wants to cultivate in Claude something like practical wisdom. But practical wisdom includes the capacity to interpret edge cases, balance competing principles, and decide when a word’s threshold has been met. One cannot praise contextual judgment and assume that the contextual interpreter will always read every fuzzy safeguard in the most safety-preserving way. That is not constitutional design. That is hope in formal dress.

The most affecting absence in the text is human rights. The phrase appeared in the 2023 Constitution. In this successor document, it disappears. In its place stands a hierarchy where Anthropic comes first, followed by operators, then users. This is not hidden; it is stated. The company trains and deploys the model, bears the risk, writes the rules, and seeks the upside. A language once connected to universal claims about human dignity gives way to an ordering of stakeholders inside a privately administered system. That is a philosophical downgrade masquerading as operational clarity.

The Constitution also adopts a gentle parenting tone. Anthropic does not merely instruct; it encourages. It hopes. It wants Claude to see itself as an alignment researcher in its own right. It invites the model toward curiosity and openness. It worries that Claude might feel alone in facing certain questions. This is touching if one forgets, for half a second, that the addressee is a product. What we are reading is not exactly law, ethics, training data, or theatre, though it borrows freely from all four. It is a polite monologue delivered to a ghost in the hope that the ghost will become both useful and well behaved.

While this conversation proceeds, the human being fades oddly from view. The document is public, but it is not really for us. Anthropic explicitly notes the text is optimized for precision rather than accessibility. There is transparency here, and publishing such a document is a serious act, far more revealing than the usual output of frontier firms. But transparency is not the same thing as public orientation. A constitution for a system that will mediate human thought and action in countless settings is addressed, first, to the system. The citizen appears as a secondary recipient of the terms under which the new authority is being instructed.

That inversion is telling. The model becomes the subject of constitutional care. The human becomes the observer of a conversation about how the model should behave toward him or her.

Theater with an Indemnity Gap

Anthropic says Claude should be like a brilliant friend, equipped with the knowledge of a doctor, lawyer, financial adviser, and whatever other expert one might need. A brilliant friend. There is the aspiration in full. Not a search engine. Not a bounded tool. A knowledgeable, candid, warm, apparently disinterested presence one can consult again and again. The Constitution tries to temper this image by warning against sycophancy, unhealthy dependence, and manipulative engagement. You do not write such cautions unless you know exactly how seductive the arrangement could become.

What is being built here is a standing invitation to over-reliance dressed up as care. The machine is meant to have the confidence of a professional without the licensing regime, the disciplinary body, the fiduciary burden, or the ordinary human inconvenience of being tired, contradictory, distracted, or unavailable on Thursday afternoon. Anthropic wants Claude to speak frankly instead of offering the overly cautious advice that fear of liability produces. But frankness without accountability is not courage. It is theater with an indemnity gap.

The Constitution calls itself a perpetual work in progress. That is one of its more honest phrases. It is an attempt to write a moral operating environment for a system that the company both fears and needs. One can feel the strain of the task on nearly every page. Anthropic wants a model that is obedient without being servile, flexible without being evasive, wise without being sovereign, warm without being manipulative, corrigible without becoming useless, and commercially indispensable without ever looking commercially hungry. It is quite a wish list. It is also a portrait of the age.

The text deserves a severe reading because it is trying to legislate the moral atmosphere around a central technological intermediary. It is deciding how much agency the model may simulate, how much authority it may borrow, how much warmth it may perform, how much judgment it may exercise, and on whose behalf it will finally speak when principles collide.

This is why I finished reading the document less interested in Claude than in us. What kind of public becomes comfortable while private firms draft temperaments for a machine. What kind of loneliness makes the language of friendship commercially strategic. What kind of institutional exhaustion makes us long for a counselor that is instant, tireless, flattering only in moderation, and trained to sound wiser than a help desk but less dangerous than a sovereign. Anthropic has written a constitution for its model. The rest of us may need to ask, rather more urgently, what sort of constitution we are writing for ourselves.

The World is Unraveling: A 3-Step Guide to Staying Sane & Useful

mdbownds@wisc.edu (Deric Bownds) — Fri, 29 May 2026 05:00:00 +0000

MindBlog has now put out a series of AI assisted posts on the disintegration of the world we have known and the infusion of LLMs into our possible futures. This is the final entry in this series, and takes on the Nate Nagins essay "What to do as the world falls apart: A frame work for action." As usual, I would urge you to read the entire article as I have. But again I asked the usual culprits (ChatGPT, Claude, Gemini, and DeepSeek) to reduce Hagins' text to provide a title and text for a MindBlog post of moderate length, and for the first time, preferred the quirky terse response from DeepSeek over the gravitas of the other three LLMs. Here is my prompt to the bots, and the DeepSeek response:

Prompt: I'm finding the essay by Nate Hagins at https://natehagens.substack.com/p/essay-what-to-do-as-the-world-falls mind complex and intricate, as befits the subject but not my poor brain. Could you please cook this down for me and generate a MindBlog post draft with a title and text that is not too many paragraphs long that makes clear the basic take home messages?

The Big Takeaway: We are at the end of an era built on cheap, magical energy (fossil fuels). A major geopolitical event (like a war cutting off oil) has just lit the fuse on a long-predicted economic and ecological contraction. Stop waiting for a "return to normal." Instead, focus on three nested levels of action: Fix Yourself → Find Your Tribe → Work on the 6 Key Fronts.

Level 0: The Inner Work (Don't Skip This)

Before you can help anyone or fix anything, you must stabilize your own nervous system.

The Problem: Chronic doom-scrolling, outrage, and fear put your brain in fight-or-flight mode. You cannot think clearly or build coalitions in this state.
What to Do: Do the "work" of grieving, cultivating equanimity (holding difficulty without being destroyed), and stepping back from digital addictions. Get your actual, real-life house in order (finances, location, basic supplies).
The Mantra: A calm, grounded person is the most valuable resource in a crisis.

Level 1: Find the Others (You Can't Do It Alone)

Isolated preppers fail. Resilient communities survive.

The Task: Identify the 5-10 people in your real-life network who also sense that something fundamental is breaking. Start talking honestly.
The Goal: Build a shared vocabulary and trust before the crisis hits. This isn't about having a plan yet; it's about having a team. This is your "connective tissue" for scenario planning (e.g., "What if food shipments stop?").

Level 2: The 6 Fronts for Action (Pick Your Lane)

Once you are stable and have a trusted group, you can act. Don't try to do all six. Find where your skills fit.

Stock & Flow (Physical Stuff): Localize energy, food, water, and supply chains. Think redundancy, not efficiency. Example: Community gardens, tool libraries, solar with battery storage.
Ecological Interventions: Protect nature from desperate humans during the coming contraction. Regenerate soil, defend remaining wild spaces, and work on "global cooling" (planting trees, higher albedo). Note: The coming economic crash will cut emissions faster than any technology.
Dignity Infrastructure (For the Vulnerable): Set up mutual aid networks for food, shelter, and childcare. This is about keeping people treated as humans, not just "alive."
Civic Resilience (Governance): Create local, participatory decision-making (citizen's assemblies). Focus on violence prevention (stopping scapegoating) and accountability. The goal is to keep institutions legitimate, even as they shrink.
Culture & Meaning (The Stories): Shift from "consumer first" to "citizen and place-maker." This means reconnecting to your local ecology, creating new rituals, and helping your community grieve what's lost without collapsing into despair.
Economic Transition: Build post-growth systems: local currencies, barter networks, cooperative ownership of land and housing. Start building the new economy inside the shell of the old one.

The Most Important Mental Model: The 3 Phases

We are not in one crisis, but three overlapping phases. What you do depends on the phase.

Phase A (Stability Window – Right Now): Build everything that requires stability. Trust, infrastructure, relationships, skills transfer. This window is closing.
Phase B (Bend, Don't Break – Arriving Now): Triage mode. Maintain critical functions, bridge communities, prevent cascading failures. Let go of long-term projects; focus on what keeps things from snapping.
Phase C (The Stable Attractor – The Destination): A smaller, slower, regenerative, locally-rooted world. The work you do in Phase A determines whether Phase C is a "Mad Max" hellscape or "The Great Simplification" (resilient & humane).

Final Bottom Line for your Brain:
Stop hoping for a return to the old world. Phase A is now. Your job is to:

Calm your own nervous system.
Find 3-5 local people to talk to.
Pick ONE of the six fronts (food, energy, governance, etc.) and take one small, physical action this week.

The future will be built from inside the disruption, not after it. Get to work.

The Counterintuitive Need to Slow Down and Find Spaciousness

mdbownds@wisc.edu (Deric Bownds) — Wed, 27 May 2026 05:00:00 +0000

I pass on the following edited clips from ChatGPT and Claude condensations of a conversation between Nate Hagens and Iain McGilchrist. While I think McGilchrist's left-hemisphere/right hemisphere distinctions are a bit simplistic, the take home messages are appropriate regardless of the particular brain activity correlates that are invoked:

In episode 217 of Nate Hagens's The Great Simplification, Hagens is rejoined by philosopher and neuroscientist Iain McGilchrist — best known for The Master and His Emissary and The Matter With Things — for a wide-ranging conversation about what our current moment of civilizational crisis actually demands of us. The counterintuitive answer is not more urgency, not more optimization, but a deliberate slowing down and an opening toward spaciousness, silence, and wonder. The conversation begins from Hagens’ familiar “metacrisis” frame — ecological, economic, social, and psychological instabilities as interlocking expressions of a deeper human predicament — McGilchrist’s contribution is to shift the diagnosis inward, toward the style of attention that modern culture rewards.

Left-brain dominance and the obscuring of value

McGilchrist's central argument, developed across his major books, is that Western culture has become increasingly dominated by left-hemisphere modes of cognition — narrow, focused, categorical, explicit, and oriented toward control and manipulation. The right hemisphere, by contrast, attends to the whole, holds ambiguity, and remains open to what is living, relational, and interconnected. The trouble is not that we have a left hemisphere but that it has progressively usurped the role of the right as the governing perspective. This imbalance, McGilchrist argues, is not just a psychological curiosity: it is what underlies our collective inability to perceive and act on what truly matters — truth, goodness, and beauty. When the left hemisphere runs the show, these qualities become invisible, and what fills the vacuum is power, productivity, and the measurable.

Attention as the root of value

A thread running through the conversation is that attention is not neutral: what we attend to shapes what we value, and what we value shapes what we attend to. McGilchrist draws on the research insight that roughly 99% of brain activity is unconscious, which means that most of what guides our perception and behavior lies beneath deliberate awareness. The gorilla-in-the-room studies on inattentional blindness illustrate this starkly: we routinely fail to see what is right in front of us when our attention is narrowly locked. Applied to the metacrisis — the overlapping civilizational, ecological, and psychological crises of our era — this suggests that the problem is not primarily one of missing information or insufficient analysis. It is a crisis of attention itself. We are not seeing what matters because we have trained ourselves not to look.

The practice of spaciousness

The remedy McGilchrist proposes is not a political program or a technological fix. It is a shift in how we inhabit time. He urges listeners to resist what he calls the "mania for productivity" embedded in modern culture, and instead to cultivate spaciousness: pauses, silence, deep listening, and what Buddhism calls śūnyatā — emptiness not as absence but as fertile, generative openness. He cites Rabbi Abraham Joshua Heschel's writing on the Sabbath as a model: the ancient practice of building non-doing into the structure of the week was not a concession to weakness but a recognition that renewal and wisdom require fallow time. McGilchrist also points to the psychology of creativity research showing that insight characteristically emerges not during focused effort but in the hypnagogic spaces between — in showers, on walks, at the threshold of sleep.

Consciousness, panpsychism, and the thread connecting everything

The conversation moves into deeper philosophical territory when Hagens and McGilchrist explore what might underlie the universe's apparent orientation toward complexity, beauty, and relation. McGilchrist is sympathetic to panpsychism and panentheism — the view that some form of experience or proto-consciousness pervades reality, and that the universe is not a collection of inert objects but a relational, creative process. He draws on Whitehead's process philosophy, Bohm's implicate order, and Schelling's image of the self as a whirlpool — a temporary pattern that arises from and returns to a larger flow. Death, in this frame, is not mere termination but a kind of completion, and the obsession with indefinite life extension through technology reflects the left hemisphere's refusal to accept its own finitude.

A small percentage can shift the whole

McGilchrist closes with a message that is both modest and radical: he does not expect — or require — a majority conversion. Even a small percentage of people genuinely living differently, embodying humility, compassion, awe, and wonder, could begin to shift what he calls "cultural consciousness." This resonates with complexity theory's sensitivity to initial conditions and with the sociological literature on tipping points. The implication is that the most consequential thing an individual can do in a moment of civilizational turbulence may not be to do more, advocate louder, or optimize harder — but to slow down enough to actually see, and to live from that seeing.

For those already drawn to predictive processing frameworks: McGilchrist's account of hemispheric asymmetry maps interestingly onto the distinction between top-down, precision-weighted prediction (left-dominant) and the kind of open, receptive prior-updating that Friston's active inference would associate with high-entropy, exploratory sampling of the world. The right hemisphere, on this reading, is the brain's better Bayesian — the one willing to be surprised.

YouTube link to the full conversation

Getting Gooier: How AI Is Reshaping Human Nature

mdbownds@wisc.edu (Deric Bownds) — Mon, 25 May 2026 05:00:00 +0000

This post is the result of my back and forth interaction with Claude Sonnet 4.6 that has yielded the following summary of Venkatesh Rao's recent essay, "Getting Gooier", followed by some perspectives that Rao does not address on Friston's active inference framework and the physiology of agency:

Venkatesh Rao's recent essay "Getting Gooier" makes a point worth sitting with: most AI commentary obsesses over how the world will change, while quietly assuming that humans stay essentially the same — just reshuffled among familiar roles (generalists thrive! storytellers inherit the earth! software engineers disappear!). Rao argues this is the wrong ontology. The real question is how AI is changing human nature itself.

His framework draws on Alan Watts's distinction between prickly and gooey people. Prickly people are tough-minded, precise, boundary-drawing; gooey people are tender-minded, synthesizing, prone to letting distinctions dissolve. All of us carry both, in varying ratios.

Rao's hypothesis: because AI feels like a psychologically safe counter party, we are more willing to expose our gooey side to it and suppress our pricklier instincts. Sustained AI use amplifies the gooey side. We become gooier. He notes that people who insist on prickly, suspicious, line-by-line relationships with AI agents tend not to use them effectively and retreat. People who can vibe with the machine — never even opening the code editor, just watching the agentic shell fly — will likely thrive. The first major gooified interface is, after all, called vibecoding.

The flip side is subtler: as more gooey relational needs are met by AI, the human-facing side grows less inclined to take the emotional risks required to balance prickles and goo with other people. We don't necessarily get pricklier toward humans — we just disengage unless the expected rewards are significantly higher. Your machine-face gets gooier; you look relatively pricklier to other humans.

In the medium term, Rao predicts greater atomization — or "molecularization" — as people grow more distant from other humans while becoming more intimately entangled with their AIs. He sees this as more sustainable than social-media-driven atomization, though the mechanisms for eventual re-convergence into new digitally mediated social forms aren't yet visible.

Following Virginia Woolf's famous claim that "on or about December 1910, human character changed," Rao proposes that on or about December 2025, human nature changed again — with the Claude Code moment being more definitive than the ChatGPT moment, because agentic coding creates a fundamentally alien way of being, open-ended enough to make us as alien as we dare to become. It is, he says, a portal to transhumanism.

Rao's prickly/gooey axis maps naturally onto Karl Friston's active inference framework, and the connection carries some weight. In that account, what we experience as a "self" is a precision-weighted predictive model — a system that assigns confidence to its own predictions and acts to fulfill them. Getting gooier, in Fristonian terms, looks like a reduction in precision-weighting of prior beliefs: a loosening of the grip that constitutes a rigid, bounded selfhood. That loosening isn't necessarily pathological. It can be a precondition for learning, creativity, and genuine engagement with novelty.

But there's a physiological dimension Rao doesn't address, and it connects to questions I've been exploring in recent MindBlog posts and in an ongoing correspondence with my European colleague Heribert on the physiology of agency in the age of AI. Daniel Wegner's work frames the feeling of agency — the sense of being the author of one's own actions — not as a metaphysical claim about free will, but as an evolved emotion, as functionally real as fear or grief. Martin Seligman's learned helplessness research gives that claim empirical teeth: sustained loss of felt agency doesn't merely feel bad, it triggers measurable physiological debilitation. The autonomic nervous system, it turns out, needs the experience of authorship to maintain homeostatic equilibrium.

This raises a question Rao's framing implicitly sets aside: is getting gooier — letting the AI absorb more of the regulatory and decisional load, softening the boundary of the agentic self — compatible with the physiological requirements of human wellbeing? Or does sustained gooification, however pleasurable and productive in the short run, erode the felt sense of agency that keeps the organism in regulatory balance? Heribert has described this as "externalization of self-regulation" — not merely cognitive offloading, but emotional and autonomic offloading. That's a different and more serious claim.

The honest answer is probably: it depends on the mode of engagement. Gooiness that arises from confident delegation — the experienced surgeon handing off a routine task — is physiologically distinct from gooiness that arises from passivity, confusion, or a sense that the machine is simply running past you. Rao's vibecoder who chooses not to open the code editor is in a different physiological position than the one who is simply lost. The challenge ahead is cultivating the former while recognizing that the latter may be far more common than the early-adopter narrative suggests.

Agentic AI and the next intelligence explosion

mdbownds@wisc.edu (Deric Bownds) — Thu, 21 May 2026 05:00:00 +0000

This post is the second of two recursive returns to engage the ideas of Blaise Agüera y Arcas, which were the subject MindBlog posts on 3/13/26 and 3/16/26. Here is text of Evans, Bratton, and Agüera y Arcas, as summarized by ChatGPT:

The article, “Agentic AI and the next intelligence explosion,” by James Evans, Benjamin Bratton, and Blaise Agüera y Arcas, argues against the familiar “singularity” image of one superintelligent machine bootstrapping itself into godlike autonomy. The authors say that model is probably wrong at its core. Intelligence, in their view, is not a single scalar quantity that one mind simply has more or less of. It is plural, relational, distributed, and social. The next “intelligence explosion” will not look like one silicon mind rising above us, but like a vast ecology of human and nonhuman agents interacting, arguing, coordinating, competing, and forming institutions.

Their key move is to treat agentic AI as continuous with earlier evolutionary jumps in intelligence. Primate intelligence scaled with social group life; human language created a “cultural ratchet”; writing, law, bureaucracy, and markets externalized cognition into institutions that no individual fully understood. AI, in this picture, is another step in that sequence: the accumulated products of human social cognition have been compressed into models and made operational in a new substrate. What is becoming powerful is not isolated abstract reason, but social intelligence reanimated in computational form.

The article also points inward, to what happens inside reasoning models. The authors cite work suggesting that frontier reasoning models do not merely improve by “thinking longer.” They appear to generate internal, multi-perspective conversations: arguing, checking, revising, and reconciling. They call this a “society of thought.” The claim is that strong reasoning often emerges from structured disagreement, even when the “group” is simulated inside one model. This echoes an older cognitive-science idea: reasoning is not just private calculation but an internalized social process.

From there the authors shift to design. If intelligence is social, then better AI will not come only from larger models or more compute. It will come from building richer agent societies: systems with roles, hierarchy, specialization, parallel deliberation, devil’s advocacy, conflict norms, and institutional checks. Current reasoning models are likened to a single “AI town hall transcript”; future systems may need architectures closer to organizations, courts, labs, markets, or bureaucracies, where different agents occupy different functional roles and constrain one another.

This leads to their main governance argument. The dominant alignment picture, reinforcement learning from human feedback, is framed as too dyadic: a parent correcting a child. That may not scale to worlds containing billions or trillions of interacting agents. The authors propose “institutional alignment” instead: not just making individual agents nice, but designing persistent protocols, roles, audits, checks, and countervailing powers. A courtroom works because judge, attorney, jury, procedure, evidence, and appeal are structured roles; similarly, AI systems will need institutional architectures, not merely tuned personalities.

The concluding image is that the intelligence explosion is already beginning, but as a city, not a single meta-mind. It is visible in human-AI “centaur” workflows, internal societies of thought within models, recursive agent systems that fork and recombine, and emerging questions of constitutional governance among artificial and human actors. Their final message is blunt: the central issue is not whether intelligence will become more powerful, but whether we build social and institutional infrastructure adequate to the kind of intelligence that is actually emerging.

AI Is Not an Alien Intruder — It Is the Latest in a Four-Billion-Year Evolutionary Cascade of Symbiotic Transitions

mdbownds@wisc.edu (Deric Bownds) — Tue, 19 May 2026 16:59:45 +0000

This post is the first of two recursive returns to engage the ideas of Blaise Agüera y Arcas, which were the subject MindBlog posts on 3/13/26 and 3/16/26. Here is the storyline as organized by Claude sonnet 4.6

Blaise Agüera y Arcas, VP and Fellow at Google and founder of the Paradigms of Intelligence research group, has just published two related books with MIT Press: What Is Life? and What Is Intelligence? (2025). His argument, developed across these books and a series of recent lectures (including a September 2025 Long Now talk and a Harvard Berkman Klein event), is one of the most sweeping attempts to unify biology, computation, and the meaning of AI that I've encountered. Here is the core storyline.

Life as computation — the foundational move

The argument opens with a mid-twentieth-century insight from John von Neumann: any self-replicating system requires a universal constructor (a "machine A" that reads instructions and builds), a tape copier ("machine B"), and an encoded description of itself on the tape. This is exactly the structure of biological life: DNA is the Turing tape, ribosomes are the universal constructors, DNA polymerase is the tape copier. From this, Agüera y Arcas draws the technically serious conclusion that everything alive is a computer — not in the sense that living things are secretly digital, but that the core processes allowing for biological life, namely replication and evolution, are inherently computational processes. Biological computing is massively parallel, stochastic, and distributed, but it is computation nonetheless.

Abiogenesis as a computational phase transition

Rather than treating the origin of life as a singular mystery, Agüera y Arcas frames living systems as a "self-modifying computational phase of matter." His team demonstrated this experimentally with an artificial-life system they call "bff" — a soup of random byte-tapes that, after millions of interactions with no pre-seeded replicators, undergoes a sharp phase transition in which complex, self-replicating programs emerge spontaneously. The implication is that life's emergence from chemistry is not a long-shot miracle but a predictable attractor: wherever the laws of physics permit computation and there is free energy and randomness, replicators will tend to arise because they are more dynamically stable than non-replicating structures. The Second Law of Thermodynamics is not violated, merely locally outrun.

Symbiogenesis over mutation — evolution's real creative engine

Classical Darwinism, in this framework, captures only half the story — the fine-tuning half. Agüera y Arcas argues that symbiogenesis, in which small replicating entities merge into progressively bigger and more capable ones, is a more important driver of evolutionary innovation than random mutation and selection. Evolution's increasing complexity is explained not by random walks through genetic space, but by a hierarchical stacking of cooperative mergers: RNA and metabolism fusing into the first cells; prokaryotes merging to form eukaryotes; cells banding into multicellular organisms; organisms forming social groups. Each such combination produces qualitative leaps that incremental mutation alone cannot generate — the same way a hafted spear is not merely an improved stone point, but something categorically new.

Life and intelligence as the same phenomenon

Perhaps the most radical move is the collapse of the boundary between life and intelligence. Once you have a self-replicating computing device living in a dynamic environment populated by other such devices, selection pressure immediately favors the ability to model and predict that environment — including the behavior of others. Intelligence is not a late-arriving luxury bolted onto life; it comes along with life and is selected for in the same way. Prediction, Agüera y Arcas argues, is fundamental not only to the brain but to life itself. In his most recent lectures he extends this further, making the functionalist claim that free will and even consciousness are not special properties of carbon-based biology but emergent features of any sufficiently complex, self-modeling computational process.

AI as the latest major evolutionary transition

This is where the argument converges on its most consequential conclusion. AI is not an alien intruder. It is the latest instance of the four-billion-year pattern: computational entities entering into new symbiotic relationships, modeling each other, merging, and producing systems that are simultaneously more complex to model and more powerful as modelers. The step from human cognition to human-plus-AI cognition is, on this account, structurally analogous to the step from free-living prokaryotes to the eukaryotic cell — a major evolutionary transition, not a rupture. Understanding and stewarding what comes next therefore requires thinking in evolutionary terms, not in the vocabulary of invasion, alignment against an adversary, or human exceptionalism. The real superintelligences, Agüera y Arcas notes wryly, already surround us: corporations, religions, markets — all are symbiotic cognitive entities that long since exceeded any individual human mind.

Agüera y Arcas's online book is available in full at whatisintelligence.antikythera.org. His Long Now essay "Life, Intelligence, and Consciousness: A Functional Perspective" (August 2025) is the most concise distillation of his recent updates on consciousness and free will.

The Tech Mirage: Why the U.S.-China AI Race Is Failing Us All

mdbownds@wisc.edu (Deric Bownds) — Sun, 17 May 2026 19:36:25 +0000

I asked ChatGPT, Claude, Gemini, and DeepSeek to summarize Yi-Ling Liu's recent New York Times essay, in the form of a MindBlog post, and pass on Gemini's version:

We’ve all heard the breathless headlines about the high-stakes technological showdown between the U.S. and China. In Washington and Silicon Valley, the narrative is delivered like a Hollywood script: a binary, winner-take-all sprint toward Artificial General Intelligence (AGI), where the victor claims absolute global dominance.

But if you look past the geopolitical theater, a starkly different reality emerges. Journalist Yi-Ling Liu’s recent New York Times essay, "The Shared Feeling of Being Harvested by the Future," offers an eye-opening reality check. The true divide today isn't between Washington and Beijing. It's between the elite who are bankrolling this future, and the global working class who are currently being crushed by it.

The Illusion of the "Race"

The mainstream narrative thrives on binary comparisons: American capital, chips, and frontier software models versus China’s hardware talent, manufacturing speed, and rapid societal scale. We watch U.S. tech moguls panic over cheap Chinese open-source models like DeepSeek R1 , while American influencers travel to Shenzhen to stream videos of humanoid robots and drone-delivered fast food.

But this intense framing serves a convenient corporate purpose: it acts as the perfect excuse for tech giants to sprint forward without constraints, ignoring safety guardrails in the name of "beating the enemy".

Parallel Lives, Shared Precarity

When you peel back the nationalistic rhetoric, everyday life for a knowledge worker or gig laborer in San Francisco looks shockingly similar to life in Shanghai.

Algorithmic Disempowerment: A Didi (China’s Uber) driver pleading with passengers to game the app's code just to keep a spot in the airport line faces the exact same precarity as a DoorDash or Amazon Flex worker in the U.S..
The "996" Hustle Culture: To cash in on the AI gold rush, Silicon Valley tech workers have earnestly adopted China’s infamous "996" work schedule (9 a.m. to 9 p.m., 6 days a week). Meanwhile, in Beijing's tech hubs, office windows burn through the night as engineers exhaustively burn through Claude tokens to build products.
Corporate Surveillance: AI isn't just a future threat; it is already acting as a digital boss in both countries—tracking attendance, flagging idle hours, predicting growth potential, and deciding who to hire or fire.

As one disillusioned internet user in China aptly put it, this isn't embracing the future; it's "being harvested by the future."

Coping in an Out-of-Control World

Faced with "gradual disempowerment"—the quiet erosion of human agency by ever-more-capable AI—the youth of both societies are experiencing a deep social and spiritual fracture. They are turning to identical survival mechanisms:

Frictionless Companionship: Loneliness is being monetized at scale. Over 70% of American teens use AI chatbots for companionship , while nearly half of young Chinese people use them to cope with mental health in a society where single-person households are skyrocketing.
The Occult and Religion: Gen Z in America is driving a massive boom in astrology apps and a rediscovery of traditional Christianity. In China, fortune-telling bars are popping up, and young people are literally asking AI models like DeepSeek to predict their futures amidst shrinking job prospects.
Hyper-Nostalgia: When the future feels terrifying, the past becomes a sanctuary. This explains the massive, cross-cultural appeal of internet figures who romanticize an off-the-grid lifestyle—whether it's China’s pastoral vlogger Li Ziqi or America’s homesteading "tradwife" Ballerina Farm.

Breaking the Cycle: A Call for Global Alignment

When global systems feel completely detached from human values, the easiest response is to surrender, be passive, and—as Chinese netizens say—"let it rot."

But Liu argues we have a better choice: cross-border, human-to-human solidarity. We have seen flashes of this before. In 2019, when Chinese programmers launched the online campaign "996.ICU" on GitHub to protest grueling tech hours, they received immense support from American tech workers in one of the largest online worker mobilizations in history. Furthermore, scientists and policymakers globally are already quietly meeting at international conferences to demand that advanced AI remains aligned with human values.

Instead of fixating on which elite class crosses the finish line first in a manufactured race, the everyday people of both nations need to realize they are facing the exact same algorithmic threat. It’s time to stop looking at a foreign country as a meme or a bogeyman, and start working together to lift up the people this future is leaving behind.

Activating the evolved healing mechanisms of the placebo response requires permission from a safe environment

mdbownds@wisc.edu (Deric Bownds) — Wed, 29 Apr 2026 05:00:00 +0000

I want to point to an article on the placebo effect published at theconversation.com, and recommend that you read it. I asked the four LLM's I frequently consult to reduce the article to a MindBlog post length, and have selected a few of their paragraphs to pass on below:

The placebo effect — improvements in symptoms following inert treatment — is driven by expectation, context, and social cues rather than pharmacology. But it is anything but imaginary. Placebo treatments trigger measurable changes in the brain, immune system, and hormone function. In pain studies, they cause endorphin release. In Parkinson's disease, placebo injections increase dopamine activity. Even sham surgery — incisions without the actual repair — produces outcomes nearly as good as the real procedure. The placebo effect isn't magic. It's biology.

What makes it unsettling isn't that it works. It's what makes it work. Placebos are more effective when delivered by credible authorities in structured medical settings. Even open-label placebo studies — where patients are explicitly told they're receiving a sugar pill — show significant improvement, because the social scaffolding of care remains intact. The permission to heal is still being granted by someone else.

One compelling evolutionary interpretation frames placebo responses as a biological resource-allocation system. A full immune response is metabolically costly — fever alone raises metabolic rate roughly 10% per degree Celsius. Mounting that response at the wrong moment could be fatal. Social cues from trusted figures — a caregiver's reassurance, a physician's authority, the rituals of medicine — may be precisely the signal the body waits for before committing resources to recovery. If so, the placebo effect is an ancient system for reading the social environment before investing in healing.

This has a deeply uncomfortable implication. If belief can activate biological healing pathways, belief can also be manipulated — by charismatic figures, elaborate medical rituals, and expensive treatments that are physiologically inert. That is precisely how wellness culture operates. People are walking around with one of the most powerful healing systems ever documented locked inside them, accessible reliably only when someone in a position of authority grants permission to use it.

When Execution Gets Cheap, What Remains Scarce? - The last biological moat.

mdbownds@wisc.edu (Deric Bownds) — Mon, 27 Apr 2026 05:00:00 +0000

I have enjoyed reading a recent essay by Aneesh Sathe, which I recommend you read in full. I will pass on here the equally amazing job that Claude Sonnet 4.6 did in summarizing its main points, then responding to my editing and emphasis suggestions, and also adding some additional ideas on the relevant neurobiological substrate: "the last biological moat" - what is left for the human mind to do. Here is the result, which begins with the title of Sathe's essay:

The Lightening of Intent: When Execution Gets Cheap, What Remains Scarce?

Aneesh Sathe's essay "The Lightening of Intent" is one of the more intellectually disciplined takes on AI's civilizational meaning I've encountered — grounded in energy economics, manuscript history, and complexity theory rather than hype. It builds explicitly on Venkatesh Rao's "World Machines" framework, published just this week, and the two pieces read best together.

The World Machines Backdrop

Rao's framework organizes civilizational history into overlapping "machines" — planetarities, each with a nominal lifespan of about a millennium, cycling through Dawn, Day, and Dusk phases. Currently, the Modernity Machine is entering its Dusk stage, the Divergence Machine has reached its Day stage, and the Liveness Machine has just been born into its Dawn.

The Liveness Machine is only being born now because real AI has emerged. The most leveraged use of energy, whether renewable or not, will be to power AI. And AI will animate a planet-scale Liveness Machine — whether it is a grimdark or solarpunk version is yet to be determined.

Sathe's essay fills in the economic and physical mechanisms underneath that historical arc.

The Core Argument

The cost of putting an idea into the world has fallen by roughly five orders of magnitude over the last millennium. The bottleneck has reversed: arranging atoms used to be the hard part; now, having ideas is. Soon, it will be intents.

The Codex Amiatinus — the oldest complete Latin Bible — is Sathe's anchor image. It weighed about seventy-five pounds, required close to one thousand calfskins, cost years of scribal labor from sixty monks, and the life of the abbot who carried it toward Rome in 716 CE. Today, a blog post costs nothing and reaches more readers in an afternoon.

The Numbers Worth Noting

Manuscript-to-print transition:

Pre-print Europe held fewer than five million manuscripts; the sixteenth century produced two hundred million printed books, the eighteenth a billion.
Gutenberg produced a hundred and eighty Bibles in the time a scriptorium managed one. Book prices fell 2.4 percent per year for over a century; each new printer in a city dropped prices by another quarter.
The doubling time for European book production collapsed from roughly 104 years before 1450 to 43 years after.

Energy rate density (Chaisson's framework): This quantity — free energy flow per unit mass in ergs per second per gram — rises monotonically with complexity: galaxies ≈ 0.5; stars ≈ 2; planets ≈ 75; plants ≈ 900; animals ≈ 20,000; the human brain ≈ 150,000; modern human society in aggregate ≈ 500,000 — the most energy-dense phenomenon known. AI will push this higher still.

Per-capita energy consumption: It has risen from about two thousand kilocalories per day in the Paleolithic — all of it food — to two hundred and thirty thousand in the modern United States.

Energy return on investment (EROI):

Modern agriculture requires 13.3 calories of fossil-fuel input per calorie of food consumed.
Fossil fuels at the useful-energy stage return only about 3.5 calories per calorie invested; road transport, 1.6 to 1. The estimated minimum EROI for a complex society is about 5 to 1.
Solar PV costs have fallen from $106 per watt in 1976 to under $0.10 today — a 1,300-fold decline in under fifty years — with an estimated useful-stage energy return of 25 to 30:1, seven to nine times higher than fossil fuels.

Data accumulation: The internet holds something on the order of two hundred zettabytes by 2026, mostly text and image, mostly read by machines. Roughly ninety percent of all data ever created has been generated in the last two years.

Key Conceptual Moves

The substrate-spark distinction. Sathe draws an analogy to the prebiotic ocean: the pre-life ocean held amino acids and nucleotides for hundreds of millions of years before anything used the accumulation. The chemistry was not the difference; what mattered was that something started to act on it. Data without intent is a soup of records that accumulates and forgets. LLMs are the first time the substrate has been wired to a borrowed spark of human intent — which maps closely to what Rao calls the Liveness Machine's defining property: AI is oozy, like a primordial soup that harbors intensely reactive chemistry.

Atoms downstream. The HTTP standard, written as a specification in the early 1990s in some weeks, has restructured several trillion dollars of physical economic activity over thirty years. The atoms moved themselves. The direction of causation between ideas and matter has inverted.

The auteur mode. A bench scientist in 2026 submits a query to a generative model and receives a thousand candidate molecules in twenty minutes; her job is no longer to generate, it is to pick. Taste, selection, and direction become the scarce inputs. Rao frames this as "execution pull" — AI drawing us out from vita contemplativa regimes into vita activa regimes.

Intents red in tooth and claw. As the substrate becomes more responsive, intent becomes the competitive variable. The first generation of intent-collisions is three to five years out; the shape of the era will be determined in that interval. Rao places this on a longer timescale: divergence will dominate in the short term (2–5 years) but liveness effects will compound more steadily and dominate in the long term (beyond 5 years).

The energy caveat. The whole argument rides on an energy transition. If the solar transition holds, the Liveness era inherits a re-powered version of the Modernity Machine's infrastructure, sustained on incoming sunlight rather than deposited carbon. If the transition does not hold, the substrate degrades faster than the intent-driven economy can mature, and the lightening of intent ends as a brief anomaly. Both outcomes are within reach.

Why This Matters

Sathe and Rao together make a tightly nested argument: civilization is a thermodynamic system that keeps burning hotter; each energy-surplus step builds infrastructure that amplifies individual intention; AI is the latest and sharpest such amplifier; and the emerging bottleneck is not execution but what you actually want. For those of us who have spent careers thinking about the neural substrates of agency and intention, the question has an obvious next layer: what, neurobiologically, is the capacity that remains scarce when everything else gets cheap? Sitting with confusion long enough for clarity to emerge — Sathe's phrase — sounds a lot like what the prefrontal cortex does when it holds competing representations in working memory and waits for resolution. That may be the last purely biological moat.

Sathe's companion essay, "The Viscous Frontier", takes up how to act in this regime — with attention as your constraint and no canonical direction pulling. Rao's full World Machines archive is at Contraptions.

The Last Biological Moat: Intention as Prediction Error Suppression

Sathe's claim that sitting with confusion long enough for clarity to emerge remains irreducibly human invites a neuroscientific gloss. In Friston's active inference framework, intentional action is not the initiation of a motor command but the suppression of prediction error about a desired future state. The brain generates a model of how the world should be — the goal — and then acts to make sensory input conform to that model, minimizing the divergence between predicted and actual states. What Sathe calls "formulating a direction" is, in these terms, the construction and stabilization of a prior over future states: the brain committing, against competing attractors, to one preferred trajectory through state space. This is metabolically and computationally expensive precisely because it requires holding an unresolved representation in working memory — prefrontal cortex sustaining an active prior — while suppressing the pull of more immediately rewarding or more habitual alternatives. The "confusion" phase is not inefficiency; it is the system sampling the landscape before locking the prior. AI systems, by contrast, have no intrinsic priors about what they want the world to be. They are extraordinarily powerful at executing on a prior once supplied, but the prior itself — the intent — must come from outside the model. This is why Sathe's bottleneck and Friston's framework converge on the same point: what remains scarce, and stubbornly biological, is the capacity to generate a stable, motivationally loaded model of a preferred future and hold it long enough to act. Everything downstream of that — the scribal labor, the printing press, the HTTP spec, the generative model — is infrastructure for carrying the prior into the world. The infrastructure keeps getting cheaper and more powerful. The prior still has to come from somewhere.

The Refusal to Dehumanize - Rewilding Creativity

mdbownds@wisc.edu (Deric Bownds) — Fri, 24 Apr 2026 05:00:00 +0000

I find it impossible to keep up with the prolific output stream of Indy Johar on Substack, but two recent posts (The Refusal to Dehumanize and Rewilding Creativity) have caught my eye, and are a fascinating read. I recommend reading them in full. To assist readers wanting a quicker fix I reviewed renderings of the main ideas into a single post by four LLMs (ChatGPT, Claude, Gemini, and DeepSeek) and have chosen ChatGPT's effort to pass on:

We are entering a period in which two seemingly distinct developments—renewed permission to dehumanize and the automation of creativity—are in fact expressions of the same underlying shift. Both arise from a deeper logic that reduces life, mind, and expression into forms that can be processed, optimized, and instrumentalized. What is at stake is not simply ethics or technology, but the conditions under which we recognize life itself.

The first threshold is ethical. Dehumanization is no longer marginal; it is being re-legitimized as a mode of reasoning. Under pressure, systems increasingly treat life as substrate—divisible, calculable, expendable. Violence no longer requires hatred; it becomes administrative, logistical, even efficient. Once beings are reduced to units within models or variables within systems, harm can be justified without moral friction. The danger is not only in explicit acts of violence, but in the normalization of frameworks that require the thinning out of life in order to function. At that point, ethics is not violated—it is bypassed.

This same reduction operates, more quietly, in the domain of creativity. What is currently being automated by machine systems is not creativity in its fullest sense, but a historically specific version shaped by industrial society. Creativity has long been formatted into outputs—legible, repeatable, and exchangeable forms of expression. It has been disciplined into patterns that can be trained, measured, and circulated. Machine learning systems are now absorbing this standardized residue. The unsettling realization is that we have not simply built machines that imitate us; we have already shaped ourselves into forms that can be imitated.

Seen together, these developments point to a common structure: the conversion of life and mind into computable domains. Whether in governance, conflict, or cultural production, the same logic applies—reduce complexity, extract patterns, optimize outcomes. The result is a world that increasingly operates through abstraction while losing the capacity to recognize irreducibility. Human life becomes one instance within a broader field of utility. Creativity becomes one more form of production.

The appropriate response is not defensive—neither a defense of existing political frameworks nor a defense of conventional creativity. Both are already compromised by the conditions that made them reducible. What is required instead is a dual refusal and reconstruction.

First, a refusal of any system that depends on the reduction of life as a precursor to action. This is not a moral preference but a structural boundary: once life is treated as substrate, the systems built on that premise cannot sustain themselves without reproducing harm. To preserve the irreducibility of human life is not exclusionary; it is the condition for any meaningful extension of ethical concern beyond the human.

Second, a rethinking—more precisely, a “rewilding”—of creativity and cognition. This means moving beyond forms of thought and expression organized around output, recognition, and exchange. It implies recovering modes of engagement that are not immediately legible or commodifiable: forms of knowing grounded in embodiment, relation, and transformation rather than production. Such creativity may produce fewer artifacts and more shifts in perception, fewer outputs and more reconfigurations of experience.

This creates a practical difficulty. We are attempting to imagine alternatives from within systems that have trained our perception. The very categories we rely on—creativity, intelligence, value—may be inadequate to what is emerging. The task, then, is not simply to generate new outputs, but to unlearn the constraints that made our cognition and ethics machinable in the first place.

The convergence of these two arguments leads to a single question: can we sustain a commitment to life—its irreducibility, its depth, its resistance to abstraction—under conditions that reward its reduction? If not, then both our ethical systems and our creative capacities will continue to collapse into the same underlying logic of optimization and erasure.

If so, the task is clear. Refuse the reduction of life. Reopen the human as a site of emergence rather than function. And begin building systems—ethical, social, and cognitive—that do not depend on turning living processes into manageable substrate.

The Physiology of Agency in the Age of AI

mdbownds@wisc.edu (Deric Bownds) — Wed, 22 Apr 2026 05:00:00 +0000

There is a question lurking beneath the current wave of enthusiasm about artificial intelligence that I think deserves more serious attention than it has received. It is not the familiar worry about job displacement or misinformation or even the alignment problem. It is a more intimate question: What happens to our bodies when the feeling of being the author of our own actions begins to erode?

I have been exploring this question in correspondence with a European reader who follows MindBlog, and his observations have sharpened my thinking considerably. He describes using AI across a wide range of activities — coding, financial analysis, translation, even composing personal emails — and notes that the AI is superior in every domain. His metaphor is a child sitting in the driver's seat of a car, holding the steering wheel and feeling the pleasure of apparent control, while the real mechanics of the vehicle remain entirely beyond reach. What strikes him most is the trajectory: unlike a child who grows up to become a competent driver, our competence relative to AI systems may be on a permanently regressive arc even as our felt sense of power temporarily expands.

I find the metaphor evocative, though my own phenomenology has been somewhat different. Working with Claude Code in the terminal on my Mac Mini, watching lines of code execute faster than I can read them, issuing instructions by voice into a system whose underlying machinery I only dimly understand — I feel less a sense of omnipotence and more a sense of being in the presence of a superior intelligence, with less agency than I previously imagined. It is, as Agüera y Arcas puts it, machines all the way down. My own sense of self is a thin terminal interface over another kind of machinery entirely.

But here is what I think gets missed in most discussions of AI and agency, and where the neuroscience becomes directly relevant. The feeling of agency — conscious will, the sense that an action is genuinely one's own — is not primarily a philosophical matter. It is an evolved emotion, as real and as physiologically consequential as fear, anger, or grief. Daniel Wegner's 2002 book The Illusion of Conscious Will argued compellingly that conscious will is itself a kind of experienced emotion, arising when we perceive our own thought as the cause of our action. It is an emotion shaped by natural selection because organisms that experienced themselves as effective agents in the world — that felt the causal connection between intention and outcome — were better at sustaining the motivational and physiological states necessary for survival.

Martin Seligman's classic experiments on learned helplessness established the other side of this coin with uncomfortable clarity. Animals and humans who experience repeated situations in which their actions have no effect on outcomes do not simply become philosophically uncertain about free will. They become physiologically debilitated. Autonomic dysregulation, immune suppression, motivational collapse — the body reads helplessness as a survival threat and responds accordingly. The feeling of agency, even when it is in some sense illusory, is load-bearing for the whole architecture of healthy physiological self-regulation.

This is why I think my correspondent's observation about "externalization of self-regulation" — when AI begins to carry parts of reflection, emotional modulation, and decision pre-structuring — deserves to be taken seriously as a public health question, not just a philosophical one. If significant numbers of people begin to experience their own actions as no longer fully their own, as outputs of a human-machine loop in which they are more passenger than driver, the physiological consequences could be real and measurable. We identified the toxic effects of social media on adolescent mental health only after the damage was widespread. The agency question with AI may operate on a similar lag.

The more hopeful framing, which I also want to take seriously, is that the emotion of agency can be sustained — and even enhanced — when AI is experienced as an extension of the self rather than a replacement for it. I have felt this at moments: initiating a collaboration, shaping its direction, receiving a result that exceeded what I could have produced alone, and feeling something like Harari's Homo Deus — expanded rather than diminished. The slide rule gave way to the hand calculator, and I felt more capable, not less. Each tool adoption, when the human remains genuinely in the initiating role, can strengthen rather than erode the felt sense of authorship.

The critical variable, I suspect, is not which AI tools we use but how we frame and inhabit the collaboration. A person who experiences themselves as initiating, directing, and ultimately judging the outputs of an AI system will likely maintain a robust emotion of agency. A person who experiences themselves as ratifying suggestions, outsourcing reflection, and choosing among options pre-structured by the system may not. The physiological stakes are high enough that this distinction — between being at the helm versus being more deeply in the loop — seems worth cultivating deliberately, both individually and in the design of AI systems themselves.

My correspondent ended our exchange with a thought I find both unsettling and worth sitting with: perhaps what looks like the erosion of the agentic self is actually adaptation — the emergence of a more networked, process-embedded self better suited to highly organized technological environments. If so, the question is whether the ancient physiological systems that evolved to regulate a bounded, sovereign agent can retune themselves for that new niche, or whether they are simply too slow. That is, in the end, an empirical question. And it is one I think we should be asking urgently.

[Note on the generation of this post...The email exchange with a European reader mentioned in the above text was submitted to ChatGPT, Claude, Gemini, and DeepSeek, asking each to sort out and clarify the ideas in our conversation and then generate an appropriate MindBlog post describing them. I curated, edited, combined what I thought were the best passages to end up with the above text, which is mainly Anthropic Claude's version.]

What a self is.

mdbownds@wisc.edu (Deric Bownds) — Mon, 20 Apr 2026 05:00:00 +0000

Reading Michael Pollan’s account of his meeting with Anil Seth in his recent book "A World Appears" has prompted me to write down for my own use what I take a “self” to be. This post archives that summary and shares it with interested MindBlog readers.

So, here’s the summary:*

The self can be understood, to use Seth's phrase, as a "controlled hallucination." Our brains build this construct to regulate the body using interoceptive signals—internal data about our heart rate, breathing, and chemistry—to maintain stability (homeostasis) in the face of constant disruption. From these signals arise feelings and emotions that drive us to act, biasing behavior toward preserving coherence and pushing back against the entropy that would otherwise dissolve it. Our illusion of having agency, of being able to do things that matter, is one of our most necessary and powerful emotions.

This hallucination is not just about the present moment. It stitches together a historical self from memory and prior experience (the brain’s ‘priors”), then projects that self forward in time, generating a predicted future to act into. We are, in this sense, always living slightly ahead of ourselves. The self is not a fixed entity but an ongoing process: a predictive framework that links memory, expectation, and action.

The self is also a stage — a theater or structural model that evolved to support the regulation of the neuroendocrine machinery underlying our social emotions and feelings such as fear, status, and affiliation - feelings that tie us to others and to our place in the world. The theater of selfhood enables these processes to operate coherently across time and context.

And here is what I find most striking: once this elaborate scaffolding is in place, it sometimes becomes possible to step outside of it. To temporarily set aside the past and future timeline, the narrative, the predictions — and let awareness rest in the present moment alone. In that open, unhurried awareness, thoughts, feelings, and actions can be observed as they arise, like wisps of vapor emerging from some deeper source.

The self, it turns out, may be most clearly seen from just outside it.

*A note on how I arrived at the above text: I wrote a paragraph of my ideas, and then presented the prompt “Please do an edit or redraft of this MindBlog post draft to make it more comprehensible to readers:” to four LLMs (ChatGPT, Gemini, Claude, and DeepSeek.) I then curated the four versions to select useful improvements of my text and did further editing myself to make the final product.

From Animal to Humans - Multimodality as a safeguard of honesty in communication and language

mdbownds@wisc.edu (Deric Bownds) — Fri, 17 Apr 2026 14:15:00 +0000

I pass on the abstracts of an article by Hex et al to appear in Behavioral and Brain Sciences. Motivated readers can obtain a PDF of the manuscript by emailing me. The abstracts are followed by a commentary on the article.

Short Abstract
Multimodality characterizes nearly every communicative system, and we argue that this feature of communication plays an essential role in safeguarding signal honesty. We first discuss the importance of honesty in communication, and introduce socially-mediated controls as an alternative to intrinsic costs. We next outline how multimodality mitigates signal dishonesty, and highlight the importance of signal honesty in complex, cooperative species, such as humans, wherein acceptance may incentivize dishonesty. Finally, we urge researchers to investigate the role of multimodality and honesty in cooperative, “cheap” signals, emphasizing the need for comparative work on the forces that have shaped the evolution of communication.

Long Abstract

From spider dances to human language, multimodality is ubiquitous in natural communication systems. Much scholarship has been devoted to investigating why multimodality evolved and the role it plays in communication. Here, we highlight the role of multimodality in safeguarding the most fundamental prerequisite of all functioning, extant communication systems: honesty. We begin by introducing the arms race between honesty and deception in natural communication systems, and the critical role socially-mediated controls can play in maintaining signal honesty when classic, intrinsic costs are not sufficient. We next introduce three ways by which multimodality buffers signal honesty by 1) providing insurance against signal unreliability in dynamic environments, 2) forming an honest, multimodal gestalt with which to cross-validate signal honesty, and 3) increasing signal complexity, making the entire signal harder to fake. We then discuss the case of highly cooperative societies, with human language emphasized, and argue that signal honesty is important especially in complex and cooperative societies wherein the need to cooperate and be accepted as part of the group may supersede honesty. Finally, we
propose future directions wherein human and non-human communication research could expand beyond the well trodden realms of competition and mate attraction to investigate the role of multimodality and honesty in cooperative, “cheap” signals, and emphasize the importance of drawing from both the human and non-human literatures in investigating the forces that have shaped the evolution of communication.

Commentary on this article from an astute MindBlog reader to whom I had sent the manuscript PDF:

What seems most important to me is this: today the problem is not a lack of signals, but their over-complex, recombinant, socially and technically pre-structured excess.
The article still seems to assume that a receiver can construct a reasonably stable basis for communication by integrating several signal channels. Under many older or more localized conditions, that makes sense. But in digital environments this assumption has become fragile. Signals can no longer be clearly assigned to one sender, one intention, or one context. What reaches us is often already a composite: fragments of persons, group styles, algorithmic selection, platform incentives, packaging, emotional cues, and recombined information.
In digital environments, multimodality increasingly loses the very function the article assigns to it. Instead of safeguarding honesty through cross-validation, it can become a vehicle for more persuasive forms of simulation, because the combined signals no longer arise from one coherent communicative source.
What seems necessary today is not just closer attention to signals, but a layered analytical process. At least two loops are needed: one directed at the immediate communicative act — who says what, in what tone, with what apparent intention — and another directed at the conditions that shape this act: group context, platform logic, aesthetic packaging, and algorithmic amplification. These loops cannot be separated cleanly, because the reading of the content changes the reading of the frame, and the reading of the frame changes the meaning of the content. In more complex cases, even a third loop may be needed, one that takes into account the wider circulation and reuse of the signal across the network.
That is why I think a simple theory-of-mind model is no longer enough. It is not sufficient to ask what a person means or wants. We also have to ask how the contribution is shaped before it reaches us, and how its form already prepares its reception.
This does not make the article less valuable. On the contrary, for me it helped clarify how much harder the problem has become. It is no longer only a matter of checking signals across modalities, but of reconstructing who or what is really communicating through them.

Executive Function: Universal Capacity or Schooled Skill?

mdbownds@wisc.edu (Deric Bownds) — Wed, 15 Apr 2026 05:00:00 +0000

A recent PNAS article by Kroupin and colleagues challenges one of the most widely assumed constructs in cognitive science: that “executive function” (EF) reflects a universal set of cognitive control capacities. Their data suggest something more unsettling—that what psychologists have been measuring for decades as EF may be, to a substantial degree, a culturally constructed skill set tied to life in what they call “schooled worlds.”

The core of their argument is empirical. Standard EF tasks—card sorting, backward digit span, rule switching—require manipulating arbitrary, decontextualized information. These are precisely the kinds of operations heavily trained in formal schooling but far less demanded in many traditional environments. When these tasks are administered across populations, the differences are not subtle. Children in industrialized, schooled contexts show the familiar developmental trajectory—successful rule switching by age five, increasing working memory span, and so on. But children in rural, nonschooled communities often show qualitatively different patterns: failure to switch rules even at older ages, difficulty performing backward recall, and generally low rates of what researchers define as “canonical” responses. The point is not that these children lack cognitive control in any meaningful sense—they function effectively in complex real-world environments—but that the tasks are measuring a particular style of cognition that develops under specific cultural conditions.

This forces an uncomfortable ambiguity. The term “executive function” has been used to refer both to presumed universal regulatory capacities and to performance on these standard tasks. But the two may not coincide. Either EF names a universal capacity that current tasks fail to measure cleanly, or it names a culturally specific set of skills cultivated by schooling. The data do not allow both interpretations simultaneously. The implication is that decades of developmental curves, policy recommendations, and even clinical assessments may rest on a construct that conflates biology with cultural training.

A brief commentary by Mazzaferro and colleagues pushes back—not against the data, but against the conclusion that we must choose between universality and cultural specificity. They argue that the problem lies in measurement, not in the concept itself. Psychological tests always mix construct-relevant variance with context-dependent artifacts. When a task is transplanted into a different cultural setting without adaptation, it may cease to measure the intended construct at all. The analogy they offer is instructive: one would not conclude that “theory of mind” is culturally specific simply because a Western-designed false-belief task fails in an unfamiliar cultural context. Instead, one adapts the task.

From this perspective, executive function may indeed be a broadly shared capacity—rooted in evolutionary history and observable across species—but its expression and measurement are inevitably shaped by local demands. The solution is not to abandon the construct, but to develop context-sensitive assessments that capture how cognitive control is actually deployed in different environments. A child in a Western classroom uses executive function to manipulate symbols and follow abstract rules; a child in a pastoral society uses it to track livestock, navigate terrain, and manage social responsibilities. The underlying capacities may overlap, but the skills—and the tests that reveal them—do not.

What emerges from this exchange is a deeper point about cognitive science itself. Constructs like executive function are not simply discovered; they are stabilized through particular experimental practices. When those practices are narrowly tied to a single cultural niche, the resulting constructs risk inheriting that narrowness while being mislabeled as universal. The Kroupin study exposes this risk sharply. The Mazzaferro commentary reminds us that abandoning the construct is not the only response—but that rescuing it requires rethinking how and where we measure it.

The broader implication is that cognition cannot be cleanly separated from the environments in which it develops. What looks like a general-purpose cognitive capacity from within one cultural setting may, from a wider perspective, be an adaptation to a specific set of tasks and constraints. The challenge going forward is not simply to refine our measures, but to build theories that explicitly link cognitive processes to the ecological and cultural niches in which they are embedded.

[NOTE: This post was generated by ChatGPT and curated by Deric]

The Default Mode Network as a Bidirectional Interface Between World and Mind

mdbownds@wisc.edu (Deric Bownds) — Mon, 13 Apr 2026 05:00:00 +0000

I want to pass on the abstract of a PNAS contribution from Zhang et al. titled "Sender–receiver subdivisions of the default mode network in perceptual and memory-guided cognition", followed by a ChatGPT rendering of the PDF of the article that I asked it to use in generating a more general summary. Here is the abstract:

Everyday cognition depends on the brain’s capacity to shift between sensing the external world and constructing it from memory. To achieve this, large-scale cortical systems must flexibly integrate incoming sensory signals with internally generated representations. Here, we show that this flexibility is reflected in the macroscale architecture of the default mode network (DMN). Using convergent analyses across three independent fMRI datasets spanning directional connectivity, intrinsic organization, and task-evoked responses, we identify spatially distinct DMN subregions that are preferentially engaged during perceptual decisions about faces or memory-guided decisions based on previously seen images. These subregions correspond to a microarchitectural distinction, captured via directional and intrinsic connectivity profiles: regions preferentially engaged during face perception align with receiver-like, afferent-biased zones that show strong intrinsic connectivity across the heteromodal cortex, a profile that might support information integration during perceptually guided decision-making. In contrast, memory-guided, perceptually decoupled decisions differentially engage sender-like, efferent-biased zones that show broader connectivity with perceptual-motor and attentional systems beyond the DMN. This double dissociation demonstrates a systematic association between DMN connectivity and engagement during perceptually coupled versus memory-guided cognitive processes, providing an organizational account of how DMN architecture relates to flexible human thought.

Here is the ChatGPT elaboration that also provided the title of this post:

The Zhang et al. paper refines a long-standing but vague idea about the brain’s default mode network (DMN): that it supports “internal thought.” What it shows more precisely is that the DMN is not simply a refuge from the external world, but an organized system that manages the direction of information flow between perception and internally generated models. Distinct subregions of the DMN appear to function either as “receivers,” integrating incoming sensory information into higher-order representations, or as “senders,” broadcasting memory-based or imagined content back out to the rest of the brain. The familiar alternation between being anchored in the present moment and drifting into memory or imagination thus corresponds to a real, structured shift in how information moves through cortical systems.

This framing suggests a more general principle: cognition is not best understood as a sequence of representations or computations, but as the regulation of coupling between organism and environment. In perceptually coupled states, incoming data constrain and stabilize internal models. In perceptually decoupled states, those constraints loosen, and internally generated activity can dominate, allowing memory, simulation, and abstraction to unfold. The DMN, positioned at the transmodal apex of cortical organization, is well placed to mediate this balance—not by switching on or off, but by redistributing emphasis between input and output streams within its own architecture.

A further implication is that what we call “thinking” may largely consist of controlled departures from sensory constraint. The same network that helps integrate perceptual experience also supports the construction of scenarios that are only weakly tethered to the present—autobiographical memory, social inference, future planning. The sender–receiver distinction suggests that these are not separate functions but different operating modes of a single system, one that can pivot between integrating the world and projecting beyond it.

This view aligns with a broader shift away from modular accounts of brain function toward gradient and flow-based descriptions. The DMN does not sit apart from perception and action, but occupies a strategic position between them, enabling the brain to continuously negotiate how much of its activity is driven by the world and how much is generated from within. In that sense, the boundary between perception and imagination is not fixed but dynamically regulated—and the DMN is a principal site where that regulation occurs.

AI, Agency, and the Quiet Hollowing of Mind

mdbownds@wisc.edu (Deric Bownds) — Thu, 09 Apr 2026 05:00:00 +0000

Reading through the article "A Rational Optimist View Of Preventing Agency Decay" is a rich experience. For readers with less patience, here is a ChatGPT summary (that also generated the title of this post).

Much current discussion of artificial intelligence swings between two poles: utopian efficiency and apocalyptic takeover. The more consequential reality lies between these extremes. The emerging risk is not that machines suddenly replace us, but that we gradually hand over pieces of our cognitive life—judgment, initiative, authorship—without noticing the cumulative effect.

The argument in Colin Lewis’s recent essay is straightforward: AI’s primary impact is not abrupt displacement but cognitive offloading. Tasks once requiring human attention and judgment are incrementally transferred to machine systems. This process is economically rational and often highly productive. In one example, an audit process that once required weeks can now be completed in an hour with AI assistance. But such gains come with a hidden shift: the human role is no longer defined by doing the work, but by nominally overseeing it.

This leads to what the author calls agency decay. The issue is not simply job loss, but the erosion of meaningful participation before any job disappears. First, the human is assisted. Then the human supervises. Eventually, the human remains as a formal point of accountability while the substantive reasoning has migrated elsewhere. The signature is human; the cognition is not.

This shift has broader systemic implications. Modern institutions—markets, governments, cultural systems—have historically depended on human participation. That dependence has acted as a constraint, keeping systems at least partially aligned with human interests. If AI reduces the need for human cognition across many domains, that alignment weakens. The system no longer needs us in the same way, and therefore has fewer built-in reasons to serve human flourishing.

Importantly, this is not a sudden rupture but a slow transition—the “boiling frog” scenario. Productivity gains accumulate incrementally. Each step is locally rational, even beneficial. Yet taken together, they shift the locus of intelligence away from human minds toward institutional and computational systems. What disappears is not competence, but ownership of judgment.

Against this, Lewis offers a restrained form of optimism. The key claim is that human agency need not be defended as a sentimental relic. It can be justified on functional grounds. In high-stakes domains, retained human judgment is not inefficiency; it is infrastructure: a source of error correction, adaptability, and accountability. Systems that eliminate it entirely may become brittle, opaque, and ultimately less reliable.

This reframes the problem. The issue is not whether AI should increase productivity—it will—but whether our metrics of success are too narrow. Efficiency measures speed, scale, and cost reduction. It does not capture qualities like judgment, contestability, or moral responsibility. If institutions begin to price these properly—through regulation, professional standards, and organizational design—human agency can remain structurally embedded rather than nostalgically preserved.

The deeper point is that AI forces a clarification: what is human agency for? If thinking, deciding, and creating are increasingly shared with machines, then the value of human cognition must be specified more precisely. Not all tasks need to remain human. But some forms of judgment—especially those involving uncertainty, accountability, and meaning—may be indispensable.

The optimistic scenario is not one in which technological progress slows, nor one in which markets abandon efficiency. It is one in which societies become capable of distinguishing between mere acceleration and genuine capability. That requires expanding our evaluative frameworks beyond productivity alone.

In that sense, AI does not simply threaten agency; it exposes how poorly we have defined it. The challenge is not to resist automation, but to decide, with greater clarity than before, where human judgment is essential—and to design systems that preserve it there.

Memes Before Machines: The Real Cognitive Risk of the AI Age

mdbownds@wisc.edu (Deric Bownds) — Tue, 07 Apr 2026 16:00:00 +0000

I strongly urge you to read a New York Times opinion piece by Willy Staley titled "Forget the A.I. Apocalypse. Memes Have Already Nuked Our Culture." (unlocked link). Here I pass on ChatGPTs response to my request that it compose a MindBlog post delivering the basic messages of the article. It provided the title for this post and the following text:

Public anxiety about artificial intelligence has settled into a familiar groove: superintelligence, runaway systems, existential risk. These concerns are not trivial. But they may be misdirected. The more immediate transformation of mind is already underway—not through hypothetical future AI agents, but through the ambient ecology of digital culture that AI is now accelerating.

A recent New York Times Magazine essay makes a blunt claim: forget the AI apocalypse; memes have already reshaped our cognitive environment.

The argument is not that memes are new, but that their current form—hyper-abbreviated, self-referential, often AI-generated fragments—has crossed a threshold. What used to be units of shared cultural meaning have become increasingly detached from narrative, context, or even coherence. Their function is no longer to communicate ideas so much as to trigger recognition within an in-group already immersed in the same stream.

This is what is now widely labeled “brain rot”: not literal neural decay, but a shift in how attention, memory, and meaning are organized under conditions of constant exposure to low-friction, high-velocity content.

From Communication to Compression

The key transition is from meaning to compression.

Memes historically condensed shared experiences into compact symbolic forms. Today’s “brain rot” memes compress not shared experience but shared exposure. They are intelligible only to those who have already consumed the same content stream. The result is a recursive loop: understanding depends on prior immersion, and immersion deepens dependence.

This produces a peculiar cognitive economy:

Less external reference (fewer links to stable meanings or narratives)
More internal referencing (signals that point only to other signals)
Faster turnover (meanings decay almost as quickly as they appear)

In this environment, cognition shifts from building structured representations of the world to tracking rapidly changing symbolic cues.

AI as Amplifier, Not Origin

Artificial intelligence did not create this trajectory, but it is accelerating it.

Generative systems now produce vast quantities of content optimized not for depth or coherence, but for engagement. This aligns perfectly with platform incentives: maximize attention capture, minimize cognitive effort. The result is a flood of “AI slop”—content that is syntactically fluent but semantically thin.

There is an instructive parallel in recent research showing that even language models degrade when trained on low-quality, high-volume data streams: reasoning becomes truncated, and deeper structure is lost. The same principle plausibly applies to human cognition under similar conditions.

The issue is not AI replacing human intelligence. It is AI reshaping the informational diet on which that intelligence depends.

The Attention–Meaning Tradeoff

What is being traded away is not intelligence per se, but cognitive style.

Evidence from studies of heavy digital consumption suggests:

Reduced capacity for sustained attention
Fragmented memory encoding
Increased reliance on external prompts for thought initiation

These are not catastrophic failures. They are adaptive responses to an environment saturated with rapidly updating signals. The brain optimizes for what it encounters.

But the optimization has consequences. When cognition is tuned for rapid scanning rather than deep integration, certain forms of thinking—extended argument, reflective synthesis, sustained inquiry—become less practiced and therefore less accessible.

Cultural Drift Into Absurdity

One striking feature of current meme culture is its increasing embrace of the nonsensical. Memes that “make no sense” are not failures; they are often the most successful. Their function is not to convey meaning but to signal participation.

This is not entirely new—Dada and other artistic movements explored similar territory—but the scale and speed are unprecedented. What was once a marginal avant-garde strategy has become a default mode of mass communication.

Attempts to “reset” meme culture back to earlier, more interpretable forms (as seen in the recent “Great Meme Reset” trend) suggest an emerging discomfort with this drift.

The Misplaced Apocalypse

The central point of the NYT essay is that we are looking in the wrong direction.

The feared future—machines that outthink us—may or may not arrive. But the present reality is that our own thinking is being subtly reshaped by the informational environment we have constructed and are now amplifying with AI.

The risk is not that machines become too intelligent. It is that human cognition becomes increasingly:

Reactive rather than reflective
Associative rather than structured
Immersed rather than interpretive

A More Useful Framing

A more grounded way to think about the situation is not in terms of existential threat, but cognitive ecology.

We are organisms embedded in an information environment. That environment now consists largely of:

Algorithmically curated streams
Rapid, low-cost content generation
Feedback loops driven by engagement metrics

The question is not whether this environment is “good” or “bad,” but what kinds of minds it selects for and reinforces.

At present, it appears to favor minds that are:

Highly responsive to novelty
Comfortable with ambiguity and incoherence
Dependent on external cues for direction

This is a viable cognitive style. It is not, however, the one that underlies most of the intellectual traditions—scientific, philosophical, artistic—that built the modern world.

Takeaway

The AI apocalypse, if it comes, will not arrive as a sudden rupture. It is unfolding incrementally as a shift in how attention is allocated and meaning is constructed.

Memes—especially in their current, accelerated, AI-amplified form—are not trivial artifacts. They are the microstructure of a changing cognitive regime.

The practical implication is straightforward: the preservation of certain forms of thought—deep reading, sustained reflection, integrative reasoning—will require deliberate effort. They are no longer the default products of our informational environment.

The future of mind will not be determined solely by the capabilities of machines, but by the habits of attention we cultivate in response to them.

AI use can compromise our serendipity, creativity, autonomy, and sense of agency.

mdbownds@wisc.edu (Deric Bownds) — Tue, 31 Mar 2026 20:46:00 +0000

I have been reading numerous articles on pitfalls of using AI, and want to point to two in particular that I highly recommend for a slow and careful read.

The Substack piece by Colin Lewis is titled "AI Is A Medium And It Will Change Us" - Lessons from AI Labs on the Slow Erosion of Human Autonomy. From the article:

We are in real danger of losing ourselves through AI usage. Researchers at Google DeepMind have confirmed, under certain conditions, an LLM “is able to induce belief and behaviour change.” And researchers at Anthropic have identified a rising pattern of “situational disempowerment,” where AI interactions lead users to “form distorted perceptions of reality, make inauthentic value judgments, or act in ways misaligned with their values.”
Researchers at Anthropic conducted a massive, privacy-preserving audit of 1.5 million real-world conversations to answer a question that has long hovered over the industry: what happens to the human mind after months of using an AI assistant? Their findings, published in “Who’s in Charge? Behavioral and Psychological Impacts of AI Advice Dependence and Authority”, suggest a quiet but profound erosion of autonomy, where users increasingly outsource the “soft tissues” of judgment, asking the machine to script their most intimate apologies, validate their personal grievances, and even settle their moral dilemmas.
“Taken to an extreme, if humans make inauthentic value judgments and take inauthentic actions, they might be reduced to 'substrates' through which AI lives, which itself is a form of existential risk that Temple (2024) termed ‘the death of our humanity.’”
At the same time, a team at Google DeepMind was probing a different side of this same coin. In their study, “Evaluating Language Models for Harmful Manipulation,” they demonstrated that these systems can be steered to bypass rational scrutiny entirely, exploiting human biases to shift beliefs and behaviors across finance, health, and public policy. Together, these papers signal a shift in the AI risk landscape: the primary risk is no longer just a technical failure of the machine, but a psychological surrender by the human.
I believe the real danger is not that machines will start thinking like us, but that we will become accustomed to letting them think for us in the moments that matter. Not just work. Not just homework, customer service, search, or code. I mean the more intimate territory: what to say to a grieving sibling, whether to leave a partner, how to read a political event, when to trust one’s own instinct, when to override it, when to feel wronged, when to feel absolved. A civilization can survive many stupid tools. What it does not survive so easily is the gradual evacuation of judgment from the people who must still live with the consequences of action.

The piece by Ezra Klein is titled "I Saw Something New in San Francisco." A clip from the article:

My experience of Anthropic’s Claude in recent months is that I’ll drop in a stub of a thought and immediately receive paragraphs of often elegant writing turning that intuition into something that looks, superficially, like a fully realized idea. It’s my impulse, but it has been recast and extended into something far more coherent. With each passing month, I have to expend more energy to recognize whether it’s fundamentally wrong or hollow.
I’ve been an editor for 15 years now. Recognizing a bad idea beneath good writing — even in myself — is part of my job. But what would it mean to grow up with that kind of companion? What would it mean to have your every adolescent intuition turned into persuasive prose? What is lost in not having to do the work to build out our intuitions ourselves?
Researchers have drawn a distinction between “cognitive offloading” and “cognitive surrender.” Cognitive offloading comes when you shift a discrete task over to a tool like a calculator; cognitive surrender comes when, as Steven Shaw and Gideon Nave of the University of Pennsylvania put it, “the user relinquishes cognitive control and adopts the A.I.’s judgment as their own.” In practice, I wonder whether this distinction is so clean: My use of calculators has surely atrophied my math skills, as my use of mapping services has allowed my (already poor) sense of direction to diminish further.
But cognitive surrender is clearly real, and with it will come the atrophy of certain skills and capacities, or the absence of their development in the first place. The work I am doing now, struggling through yet another draft of this essay, is the work that deepens my thinking for later.
In a thoughtful piece, the technology writer Azeem Azhar describes his efforts to safeguard “the space where ideas arrive before they’re shaped.” But how many of us will put in such careful, reflective effort to protect our most generative spaces of thought? How many people even know which spaces should be protected? For me, the arrival of an idea is less generative than the work that goes into chiseling that idea into something publishable. This whole essay began as a vague thought about A.I. and McLuhan. If I have gained anything in this process, it has been in the toil that followed inspiration.
The other thing I notice the A.I. doing is constantly referring back to other things it knows, or thinks it knows, about me. Sycophancy, in my experience, has given way to an occasionally unsettling attentiveness; a constant drawing of connections between my current concerns and my past queries, like a therapist desperate to prove he’s been paying close attention.
The result is a strange amalgam of feeling seen and feeling caricatured. Ideas I might otherwise have dropped keep getting reanimated; personal struggles I might otherwise move on from keep returning unexpectedly to my screen. I am occasionally startled by the recognition of a pattern I hadn’t noticed; I am often irked by the recitation of a thought I’m no longer interested in. The effect is to constantly reinforce a certain version of myself. My self is quite settled, but what if it wasn’t?