<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Lucas Vollet, PhD on Medium]]></title>
        <description><![CDATA[Stories by Lucas Vollet, PhD on Medium]]></description>
        <link>https://medium.com/@lucasvollet?source=rss-e184748bf401------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*C1I-sq8gEdhud2SXSea8Gg.png</url>
            <title>Stories by Lucas Vollet, PhD on Medium</title>
            <link>https://medium.com/@lucasvollet?source=rss-e184748bf401------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Wed, 18 Nov 2026 11:44:33 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@lucasvollet/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Machine-Done, Machinable: A Rule to Start Judging Human Uniqueness Again.]]></title>
            <link>https://medium.com/@lucasvollet/machine-done-machinable-a-rule-to-start-judging-human-uniqueness-again-ef960074f0d5?source=rss-e184748bf401------2</link>
            <guid isPermaLink="false">https://medium.com/p/ef960074f0d5</guid>
            <category><![CDATA[llm]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[writing]]></category>
            <category><![CDATA[ai-education]]></category>
            <category><![CDATA[education]]></category>
            <dc:creator><![CDATA[Lucas Vollet, PhD]]></dc:creator>
            <pubDate>Sat, 24 Oct 2026 10:55:33 GMT</pubDate>
            <atom:updated>2026-10-24T10:55:33.081Z</atom:updated>
            <content:encoded><![CDATA[<p>Let’s propose a rule for dealing with a question that has become increasingly difficult: how can we judge human uniqueness now that so many of the signals and proxies we conventionally relied upon have collapsed? Is it still possible to conduct a <em>fine-toothed examination</em> of human contributions without simply reproducing our old prejudices about authorship, talent, and creativity? Well, remember that one of the seemingly trivial moves that made Descartes and Husserl famous was the radicalization of doubt. Husserl called his version the <em>epoché</em>: putting certain assumptions in parentheses, suspending their authority in order to investigate what remains available to judgment.</p><p>Why shouldn’t we try something similar?</p><p>The rule I find most natural is this: whatever the machine does is MACHINABLE<strong>.</strong> It sounds trivial? Wait.</p><p>Whatever the machine can do, consider it mechanical. If it can produce a beautiful metaphor, a convincing argument, an elegant translation, or an ingenious joke, then those things are mechanically reproducible. We cannot subsequently retreat into the belief that these accomplishments must contain something sacred, irreducible, or intrinsically incompatible with mechanical production.</p><p>For the purposes of this experiment, let’s even adopt a deliberately crude equation: mechanical means cheap. This equation is certainly debatable, but that is not our concern yet. We are establishing a methodological restriction, not announcing a metaphysical discovery.</p><p>Now, why this rule is so indispensable for us to begin this debate? What irrational reactions it blocks? Lets see.</p><p>Under our provisional rule, we have deliberately removed the option of attributing those exceptional qualities to mechanical production. Here, again, we must be consistent with our initial concession: if something is done by a machine, it is machinable.</p><p>This may sound outrageous to someone who cannot find, within the accomplishments of LLMs, anything that distinguishes them from the finest achievements of human intelligence. But it is also remarkably liberating. Once we finally come to terms with our assumption, we find ourselves in an interesting position: IF — and whenever — we are certain that we are witnessing something good, something genuinely interesting — whether a powerful argument, a beautiful analogy, or a moving work of art — and yet we have already ruled out the possibility that its significance lies in the mechanical contribution itself, then we have to give it to HUMANS.</p><p>If both commitments are maintained — AND THAT IS WHAT THE RULE GIVES US — the conclusion becomes unavoidable: whatever makes the achievement meaningful must come from the human side of the equation.</p><p>But this also makes something else necessary, and it is perhaps the most liberating consequence of the rule.</p><p>If you suspect that something is machine-produced, mechanical, and trivial, then you already have your answer. No Pangram detector is required. Whatever features of the work led you to that judgment — whatever standards you applied, consciously or unconsciously — have already told you that you consider the text machinable. And this judgment would remain exactly the same even if the text had been produced without a single shred of machine assistance, entirely by hand, through the most painstaking artisanal labor.</p><p>So my rule is liberating on two fronts.</p><p>First, if you find something beautiful, original, or meaningful, then, under our assumption, whatever makes it so must come from the human side of the equation. Now find out what that contribution is.</p><p>Second, if you find something trivial, formulaic, or machinable, you already have your verdict. Who cares whether a machine actually produced it? It could have been written entirely by hand. No AI detector required.</p><p>And finally, a third liberation: the rule weakens the fear that using machines somehow makes our human abilities to rot.</p><p>I know this is controversial. If a machine starts producing my analogies, isn’t the part of my brain responsible for making analogies slowly rotting away?</p><p>Well, our rule says: if the machine can do it, it is machinable, so you are losing nothing but a mechanical thing you used to do with pride.</p><p>This is a serious blow against studies to measure cognitive offloading: to defend the offloading thesis as a loss of something intellectually essential, one must first reject our rule. Many studies on AI and cognitive offloading, including the much-discussed MIT research, risk assuming precisely what is under investigation: that the abilities being outsourced are the very abilities we ought to preserve as central to human intelligence.</p><p>Then they measure whether people become worse at performing those tasks without assistance and discover — of course — that outsourcing a task reduces our independent practice of it.</p><p>But what has that established? That we are losing a skill, certainly. Not that we are losing something intellectually essential.</p><p>Remove the assumption that the mechanically reproducible skill is itself the seat of intellectual value, and these studies lose their immediate force.</p><p>Objections:</p><p>Now, someone might object that this is suspiciously easy. Haven’t we simply invented a convenient route back to human priority? Haven’t we guaranteed our own victory by definition?</p><p>But consider what the objection would require. Our conclusion follows from a deliberately trivial rule: what is machine-done is machinable. Unless we are prepared to admit that the machinable, mechanical components of an achievement can themselves be the bearers of its meaning, the supposedly easy road turns out to be the only road consistent with our assumptions.</p><p>And if we do want to challenge that conclusion, we must challenge the assumption openly, rather than preserving the prestige of mechanical accomplishments while denying it to the machine. The rule BLOCKS that contradiction. So, its usefullness is not that we know it to be true (it may urn out to be false: that some of the machine’s contribution belong to the meaningfull parts of it), but that it BLOCKS the contradiction that is blocking people’s brain in this debate: the assumption that the machine’s achievement are mechanical, trivial, parrot-like and at the same time can cheat human judgment, be above human rationality, imitate and fool even our best feelings and the ceiling of academic-peer-review and art-critic. My rule takes us out of this PERVASIVE confusion. So, if anything, it has to be adopted to even start moving the debate out of its first confusion.</p><p>Now, suppose the rule is indeed correct. What is left? The real heavy lifting is now something closer to <em>ego hygiene</em>.</p><p>We must clean up the wounds left by discovering that some of the things we regarded as exceptional in our own talents were, after all, machinable. Everyone has something to lose here. The writer discovers that elegant prose is machinable. The artist discovers that striking compositions are machinable. The philosopher discovers that sophisticated analogies, objections, and even apparently original arguments are machinable. Whatever still is exceptional, is there — our hearts KNOW — but they arent those old things anymore. So: clean ego. It hurst but we have no option.</p><p>We have to learn to say, without resentment: “Well, something I considered exceptional about my abilities turned out to be mechanically reproducible.”</p><p>And then comes the much more difficult task: judging originality, significance, and meaning for their own sake, without relying on the mechanical proxies that once made those judgments so comfortable.</p><p>Because under our rule, if something is mechanically reproducible, that reproducibility cannot be what makes it meaningful. We must look elsewhere.</p><p>This is where the investigation of human uniqueness can finally begin again.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=ef960074f0d5" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Is It Really So Obvious That AI Is Corrupting Education?]]></title>
            <link>https://medium.com/@lucasvollet/is-it-really-so-obvious-that-ai-is-corrupting-education-4e37361a5aba?source=rss-e184748bf401------2</link>
            <guid isPermaLink="false">https://medium.com/p/4e37361a5aba</guid>
            <category><![CDATA[ai-writing]]></category>
            <category><![CDATA[ai-education]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[education]]></category>
            <dc:creator><![CDATA[Lucas Vollet, PhD]]></dc:creator>
            <pubDate>Thu, 27 Aug 2026 09:57:11 GMT</pubDate>
            <atom:updated>2026-08-27T09:57:11.196Z</atom:updated>
            <content:encoded><![CDATA[<p>Billionaires and water consumption are easy targets for pseudo-arguments and distractions from the real problem in education. Take that away, and we immediatly realize that calling something “cheating” because it threatens a hegemonic and deeply elitist system of education is not an argument: it is a deffense mechanism that distract from the underlying epistemology.</p><p>When examining the literature — free from the distractions that drive people to rail against AI as if it were a new civic duty — one encounters a vast landscape lacking consensus, comprising diverse traditions and respected authors who, if viewed through the lens of their background theory, would not oppose the way AI is restructuring the educational system. The way AI mediates cognition and negotiates meaning with an interlocutor is directly aligned with active externalism (Clark &amp; Chalmers) and semantic pragmatism (Wittgenstein, Quine): social, negotiated practice rather than an internal mental possession. Vygotsky’s <em>Zone of Proximal Development</em> posits that learning occurs best through mediated, social dialogue with a “More Knowledgeable Other.” One may, of course, deny the machine the role of the “Other,” but one cannot deny the value of its mediation without lapsing into deep denialism.</p><p>This connects to other figures deeply involved in leftist educational thought, like Paulo Freire — for whom education is never neutral, but an arena of power, access, and dialogue. AI disrupts the hegemony of a dominant vocabulary by allowing the reconstruction of angles and gradients of understanding directly within the loop of question and response. In this sense, it destabilizes power structures in education rather than flattening them into uniformity (this, again, despite the billionaires).</p><p>This sums up as deeply inconsistent with the mythology of the isolated “gifted mind” that much of today’s anti-AI advocates invoke when they claim one should not do art if not with their hands, and it is also inconsistent with the normative myth of “normal learning” that much of today’s anti-AI advoates invoke to claim one should not write unless if he learns it as a “normal person” — a mythology that has always been a proxy for the mind that already owns the dominant vocabulary.</p><p>Even if we grant that AI tends to standardize language, practices developed through its assistance can still absorb and accommodate a remarkable diversity of linguistic and cognitive styles. That is why AI is becoming a kind of barricade for neurodivergent people and non-native speakers: dialogue with the machine allows them to build meaning through successive layers of revision, clarification, and reformulation, rather than forcing them into the linear game of fluency, speed, and mastery of a dominant vocabulary.</p><p>Finally, Even amid attempts to deny that the machine is doing what it clearly does — dressing it up as some kind of “decoy” — I rest on a simple point: on the spectrum of positions in this debate, we’re already tilted far more toward AI than away from it. I rest on this because, different from the hysteria’s image of previous educational system as some metaphysical Default, I am not selling the myth that AI is perfect. I am saying the aggressive hostility and moral entitlement displayed by these critics are wildly disproportionate to the actual state of the debate — they’ve broken free of any respectable orbit of intellectual rigor and drifted into polarizing, even dangerous, territory. That shift puts traditional education systems on the defensive; they don’t deserve to be treated as some untouchable, divinely-ordained default. And it certainly doesn’t justify the frenzied hysteria, nor the rude and aggressive way the public grants itself permission to shame AI assistance.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*66L9cWQZ1gO3WNvMA9uL7w.png" /></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=4e37361a5aba" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[It Doesn’t Think — But It Stole Your Thinking? (The Price of a Side).]]></title>
            <link>https://medium.com/@lucasvollet/it-doesnt-think-but-it-stole-your-thinking-the-price-of-a-side-b07dfbef997b?source=rss-e184748bf401------2</link>
            <guid isPermaLink="false">https://medium.com/p/b07dfbef997b</guid>
            <category><![CDATA[can-ai-think]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[ai-thinking]]></category>
            <category><![CDATA[ai-writing]]></category>
            <dc:creator><![CDATA[Lucas Vollet, PhD]]></dc:creator>
            <pubDate>Mon, 24 Aug 2026 13:41:18 GMT</pubDate>
            <atom:updated>2026-08-24T13:41:18.017Z</atom:updated>
            <content:encoded><![CDATA[<blockquote>Why denying machine thought forces us to redefine thinking — or abandon the myth of the self-sufficient human author</blockquote><p>Among the most persistent contradictions operating beneath the panic about AI is this one: artificial intelligence is said not to think, and yet it is also said to think <em>for</em> you, to rob you of something, to have taken over an activity that was supposed to be yours. Both claims get asserted, often in the same paragraph, often by the same person, with no sense that anything needs reconciling.</p><p>The reason nobody reconciles them is that each claim, on its own, feels earned. Someone feels like a machine cant think. But he also watches a paragraph or paper on the difference between democracy and tyranny fall out of a prompt in four seconds. So he feels like both things are true: it is not thinking, but it is preventing you to think. Both things can’t be true.</p><p>We have to talk deeper about the psychology behind this: when someone watches an entire argument about democracy and tyranny appear after a prompt something in them registers a violation — as if a small sacrilege had occurred. The paragraph was supposed to cost something. It was supposed to be the visible residue of years of reading, of an education, of the specific labor that makes a person capable of saying a true thing well. Now the residue exists without the labor, and the discomfort that follows gets described, understandably, as “the machine is thinking for me.” Here we already have a problem : a person observing the same fact could say: maybe that paragraph about tyranny was never something sacred, it can be automated so, yes, it is not even something that should require that whole work of education and etc. So seeing this as a problem is already a choice, not a necessity. People avoid that conclusion, of course, not because it is false, but because it implies something our era is simply not prepared to grant: that much of what constitutes the visible output of academic life may simply be cheap to produce. It may carry nothing transcendental, uniquely human, or intellectually exceptional.</p><p>So, the contradiction is not inevitable. It is, in fact, trivially escapable: admit the second reading, and the tension dissolves on the spot. There was never a law of thought forcing anyone to hold “it doesn’t think” and “it stole my thinking” together. What forces it is something closer to refusal and denial — and refusal, unlike logic, can be walked back.</p><p>But watching that refusal get walked back is instructive, because it does not happen in one step. The first move, when someone is pushed to choose between the two readings, is something like anger — the recognizable anger of a person being asked to give up a story they need: the narrative that, to write a paper or reason about democracy, one needs a vast amount of intellectual work (and money). What is being defended in that anger is not a belief but a position: the sense that one’s formation was not merely useful but exceptional, irreplaceable, close to sacred. Push past that anger, though — and some people do get pushed past it, because there is no way to hold the first reading without attributing denial — and what is left is not resolution. It is a second kind of confusion, and this one is not made of ego at all. The deeper issue, then, is not merely ego. It is a crisis of available explanation.</p><p>The second confusion is that we do not yet have a satisfying account, in terms anyone would call scientific, of what has actually happened, when we see the things this machine can do. Rather than concede that the machine did the thing — produced a real paragraph doing real explanatory work on democracy and tyranny — some people preserve the first pole of the contradiction (“it doesn’t think”) by denying the output any substance at all. The text is redescribed not as text but as a kind of decoy. It didn’t <em>teach</em> anything; it <em>mimicked</em> teaching, the way a lure mimics a minnow. And once that redescription is in place, a second problem appears immediately, because something is still visibly happening — people are reading these paragraphs, being moved by them, they can make a debate work in the same way something written by human hands. If the text has no real content, that effect needs a different explanation, and the explanation on offer is not displacement but infiltration. This is the move underneath Dead Internet theory in its more totalizing forms: the machine’s output isn’t valueless prose that happens to be inert, it’s active decoy, closer to a spy or a parasite than to a text — something whose entire function is deception, occupation, the hijacking of attention and trust, rather than the honest performance of an intellectual task.</p><p>The appeal of this move is that it lets a person hold onto both halves of the original contradiction without ever resolving the tension, because it relocates the tension outside the machine and into a conspiracy. “It doesn’t think” is preserved intact — nothing in this account has to grant the machine competence. And “it’s stopping us from thinking” is preserved too, but reinterpreted: not as competition (a better producer of the same good), but as sabotage (an agent working, in something like bad faith, to disable the human capacity rather than to out-produce it). But this contradiction, like the others, cannot be held for long, and for a very concrete reason: even granting that the machine is nothing but a decoy dressed up as prose, no one with their reasoning intact can stand in front of a student and deny that the paragraph on the origins of democracy is correct, once they have read it and confirmed that it is. That option simply isn’t available to anyone who wants to keep their own credibility. You cannot sit there as a competent adult and say “I refuse to grant that this is a sound, well-written account of the origins of democracy” while privately knowing that it is sound, and about as well written as most humans manage — though naturally a few exceptional ones would still do better.</p><p>Yet more than professional ego is blocking the debate. We do not possess a stable and publicly shared account of what has happened. There is, of course, extensive work on distributed cognition, extended minds, the collaborative production of knowledge, structural linguistics, and the dependence of individual thought on inherited symbolic systems. Readers of Putnam, Vygotsky, structuralism, or the literature on situated cognition were therefore somewhat better prepared to understand intelligence as mediated rather than sealed inside an autonomous author. Someone who had absorbed those traditions had, at least, a conceptual seat to sit in — a way of already suspecting that “a person’s thought” was never as self-contained as the romantic picture assumed, so the shock of a machine producing competent argument was a difference of degree, not an ontological rupture. But that was never more than a small fraction of the people confronting this. Almost no one, in 2022, had a framework ready for the actual event: a mechanical process generating, on demand, a coherent paragraph teaching roughly what nine in ten competent teachers could teach about democracy and tyranny. That is not a small update to prior belief. It is an event without a genre.</p><p>This was an extreme epistemic shock.</p><h3>Final Chapter: The Price of a Side</h3><p>Return, then, to the shock itself, because it is real and it deserves to be taken as seriously as the confusion it produces. Almost no one alive today had a framework ready for a machine producing, on demand, sound and well-formed argument about things that used to require years to argue well. Given that shock, it is understandable that most people find no way out and simply learn to live inside the unresolved contradiction, treating it as a permanent condition rather than a temporary and correctable state. But living inside a contradiction is not a cost-free arrangement, however tempting it looks from the outside. A person who holds “it doesn’t think” and “it’s stealing my thought” without ever settling which one is doing the real work loses, along with the contradiction, the ability to say anything precise about what they are afraid of. Their fear stops being an argument and becomes a mood — loud, persistent, and strictly uninformative. Most of the hysteria surrounding this subject is, in this sense, not a coherent fear at all. It is the noise a mind makes when it refuses to pick a side.</p><p>So a side has to be picked, and I will say which one I pick, while granting in advance that I am not certain of it. I don’t think the machine thinks. But saying that honestly, rather than as a slogan, turns out to cost something. To hold “it doesn’t think” against everything the last few years have produced, you cannot let “thinking” mean what it casually meant before — competent synthesis, correct argument, fluent prose on a hard topic — because the machine now does all of that reliably. You have to narrow the word until it names something rarer: not the ability to produce a sound paragraph on democracy and tyranny, which is now cheap, but the ability to generate a genuinely unprecedented angle on a debate — the kind of move a person makes who wakes up daily with new, uncoerced ways of putting Kant and Hume in the same room, cross-theoretical operations that did not exist as options before that mind produced them. I know someone like this, and I believe him when he says that no amount of machine assistance would have handed him those angles on its own; the machine can dress the sentences, never supply the move. Choosing to believe this saves “the machine doesn’t think,” but it saves it at a real price: it restricts “thinking,” honestly used, to a small, gifted class of people, so of course the price here is high.</p><p>The other option available to the same camp goes the opposite way. Take the externalist and structuralist literatures seriously — the idea that thought was never purely internal, that knowledge has long been distributed across people, tools, and institutions rather than housed whole inside one skull — and the machine does not need an interior to do real work in that economy. It can hold a position steady that a person would otherwise lose, recover an argument’s better footing when a person’s own drafting collapses under fatigue or haste, and shift a person’s stance a little further along the gradient toward the more defensible view — a kind of gravitational assist, useful precisely because the pull runs toward truth rather than away from it, which is the only direction such an assist could run without being self-defeating.</p><p>On this account “the machine doesn’t think” survives too, but survives by dissolving the very thing that made thinking feel special in the first place: the picture of the individual, self-sufficient thinker whose paycheck and status depended on that picture being true. The price here is not that high, but it is high for academic gatekeepers earning money from that myth so we may expect push back from them.</p><p>Both routes let you keep the sentence “the machine doesn’t think.” Neither lets you keep it for free. One buys it with an elitism about who counts as an original mind; the other buys it by handing the last few centuries of romantic authorship over to a distributed, social, already-collaborative account of cognition that was arguably always more accurate and always less flattering.</p><p>I don’t see a way to hold the sentence without paying one of these two prices, and I don’t think the debate gets to keep pretending otherwise for much longer. Eventually, one way or another, a choice gets made — and the sooner it is made explicitly, rather than smuggled in under the cover of an unexamined word, the sooner the debate stops mistaking its own confusion for an argument.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*QauHlvbhA4ix7EswTCUvRg.png" /></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=b07dfbef997b" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Yuval Noah Harari’s AI Argument Is a Symptom, Not a Diagnosis]]></title>
            <link>https://medium.com/@lucasvollet/yuval-noah-hararis-ai-argument-is-a-symptom-not-a-diagnosis-05ec29956909?source=rss-e184748bf401------2</link>
            <guid isPermaLink="false">https://medium.com/p/05ec29956909</guid>
            <category><![CDATA[yuval-noah-harari]]></category>
            <category><![CDATA[ai-agent]]></category>
            <category><![CDATA[ebook-yuval-noah-harari]]></category>
            <category><![CDATA[ai]]></category>
            <dc:creator><![CDATA[Lucas Vollet, PhD]]></dc:creator>
            <pubDate>Fri, 21 Aug 2026 20:37:19 GMT</pubDate>
            <atom:updated>2026-08-21T20:37:19.069Z</atom:updated>
            <content:encoded><![CDATA[<p>Before I say anything else, watch the video bellow (link in the end). Harari is a serious thinker, and whatever you make of the argument, it’s worth encountering in his own words — the historical range is real, the rhetorical craft is real, and the anxiety he’s articulating is one a lot of smart people share right now.</p><p>That said, I want to lay out where I think it goes wrong, and why I think it matters beyond just “one thinker’s take.” Because what Harari is doing here isn’t an isolated argument — it’s a symptom. It’s part of a broader pattern I’d call AI hysteria: a mode of talking about these systems that borrows the vocabulary of rigorous analysis while actually functioning as something closer to myth-making. And it sits on a spectrum that, at its far end, shades into full conspiratorial thinking — the “dead internet theory” crowd, the sense that nothing is what it seems anymore, that agency has been secretly transferred to machines while we weren’t looking.</p><p>This is what I find so difficult to watch being presented as a coherent position when it never quite manages to become one. Harari’s argument contains real insights, considerable rhetorical power, and obvious familiarity with the relevant historical problems. But its argumentative strategies repeatedly cannibalize one another. Instead of converging on a clear thesis capable of surviving revision, they produce a salad of half-articulated fears, behavioral observations, and apocalyptic metaphors. At the center is a contradiction shared by much AI hysteria: Harari combines the fear that machines will take over the world with an underlying contempt for machines and for everything classified as “merely mechanical/BUREOCRATIC.”</p><p>Thinkers like Harari deeply invested in language-first theories — accounts that reduce reason to the operation of historically dominant vocabularies — have no easy way out. If reason was always merely the mask worn by a ruling narrative, then AI does not overthrow an autonomous rational faculty. It simply reflects this linguistic machinery back at us and accelerates its repetitions.</p><p>Yet Harari abruptly attributes extraordinary powers to that same machinery: it will travel to the stars, take control of civilization, manufacture intimacy, and deceive us through dead-internet populations of artificial humans. But why should deception matter if reason was never anything more than submission to a dominant vocabulary? This may look like an innocent inconsistency, but I think it is the central confusion that makes his discourse feel like a dead end.</p><p>Harari has no clear account of human thought or rationality as a capacity that stands above, evaluates, or resists linguistic and bureaucratic machinery. Yet he retains the dramatic imagery of a war between humanity and the machine. Such a war makes no sense under his own premises. If human beings were always creatures governed by stories, codes, bureaucracies, and mechanical systems of coordination, then AI is not an alien enemy confronting an independent human reason. It is only another stage of the machinery that already constituted us.</p><p>Harari therefore needs humanity to be thoroughly mechanical in order to diminish its pretensions, but genuinely rational and autonomous in order to make its coming defeat by AI tragic and dramatic. In some of his videos, he even warns of “the collapse of truth,” while operating within a framework that leaves no clear space for truth independent of language, narrative, and historically dominant vocabularies. But if truth was always only an effect of the stories organizing society, what exactly is now collapsing?</p><p>What frustrates me in this bunddle is not that he refuses to choose my side. It is that he refuses to choose any side. The result is a fabricated hysteria staged within a conceptual framework that has already eliminated the possibility of a solution. He flatters AI panic without granting that there is anything left to panic about, because by his own standards we lost our agency long ago, when the first cultures began “outsourcing” it to language, mythology, bureaucracy, law, and other systems of collective coordination.</p><p><a href="https://www.youtube.com/watch?v=_V_ed5fuexA">https://www.youtube.com/watch?v=_V_ed5fuexA</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=05ec29956909" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The Prosthesis of Suspicion]]></title>
            <link>https://medium.com/@lucasvollet/the-prosthesis-of-suspicion-39638c1f16be?source=rss-e184748bf401------2</link>
            <guid isPermaLink="false">https://medium.com/p/39638c1f16be</guid>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[how-ai-detectors-work]]></category>
            <category><![CDATA[neurodiversity]]></category>
            <category><![CDATA[fraud-detection]]></category>
            <dc:creator><![CDATA[Lucas Vollet, PhD]]></dc:creator>
            <pubDate>Fri, 24 Jul 2026 14:28:17 GMT</pubDate>
            <atom:updated>2026-07-24T14:28:17.994Z</atom:updated>
            <content:encoded><![CDATA[<p>About AI detectors and wha they really track</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*CtepgMSzKj3wkqM_81pdJw.png" /></figure><p>Some theorists working at the intersection of evolutionary psychology, anthropology, and cognitive science have suggested that human beings evolved to track differences such as skin color because such cues once served as biologically useful signals requiring social reinforcement. On this picture, distinguishing insiders from outsiders helped protect tribes from impostors and coordinate cooperation.</p><p>I have always found this an astonishingly low opinion of our biological instruments. But perhaps evolution is under no obligation to preserve our dignity. Whatever strategies happened to increase survival were the ones that stayed.</p><p>The real philosophical question is different. What exactly was there to be tracked? And how much confusion had to masquerade as knowledge for that biological need to be satisfied?</p><p>Today we have AI detectors. Unlike our perceptual systems, these are prosthetic devices — artificial classification machines built to identify another distinction that many people desperately want to exist. Their task is even stranger than anything our brains were supposedly selected to do: to slice a text into percentages of “human” and “machine,” as though the countless routes through which a person acquires language, concepts, theories, memories, biographies, second languages, conversations, books, teachers, and now computational tools could be decomposed into a numerical ownership structure.</p><p>Twenty-five percent machine.</p><p>Fifty-five percent human.</p><p>Twenty percent collaboration.</p><p>Presented as though reality itself came pre-divided along those joints.</p><p>But the joint is not there.</p><p>There is no principled boundary separating “the human part” of a sentence from “the machine part”. There is only the convention used by a machine to correlate metaphysical fantasy and fetiche with numbers. Was a concept learned from a teacher more human than one refined through a book? Is an argument polished by an editor more authentic than one clarified through translation software? Does thinking in a second language reduce one’s humanity? These percentages pretend that there exists a hidden ontology waiting to be measured, when in fact they merely reflect the output of a statistical classifier searching for patterns it was trained to recognize.</p><p>What is remarkable is that a major intellectual platform would institutionalize such a category.</p><p>One cannot avoid wondering what social desire this device is really serving.</p><p>The desire seems obvious enough: to preserve a symbolic hierarchy in which certain stylistic habits — once monopolized by a relatively homogeneous educational elite — could continue functioning as badges of superiority. As computational tools erode those accidental advantages, a new ritual becomes necessary. If authors can no longer be excluded by fluency alone, perhaps they can be ranked by the supposed percentage of “machine” contaminating their prose.</p><p>The detector is therefore not merely a piece of software. It is a prosthesis for a distinction that no human cognitive system ever naturally possessed. It extends not perception but a social fantasy.</p><p>Ironically, even for those who welcomed it as an instrument of humiliation of the “inferior” (those needing machine-assistace), it performs rather poorly. There is little emotional satisfaction in denouncing someone as “twenty-five percent fraudulent.” The moral theater collapses under its own absurdity. The accusation lacks the certainty required to produce the public degradation its users seem to seek.</p><p>And that leaves an even more practical question.</p><p>What does this episode say about Substack itself?</p><p>If a platform of this size genuinely believes that displaying arbitrary percentages of “human” and “machine” provides meaningful information about authorship, writing, or intellectual contribution, or even worst, about “humanity” — as if there is one single type of brain deserving “Humanity-status”- one begins to wonder about its capacity to understand the very community it hosts. More importantly, one wonders how well it is positioned to compete with platforms like Patreon or even YouTube.</p><p>The detector does not reveal a hidden structure of writing.</p><p>It reveals a hidden anxiety about the democratization of intellectual production.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=39638c1f16be" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The Carnapian Cop-Out: Why Anti-AI Anxiety is Just Bad Philosophy]]></title>
            <link>https://medium.com/@lucasvollet/the-carnapian-cop-out-why-anti-ai-anxiety-is-just-bad-philosophy-78aba8413bb6?source=rss-e184748bf401------2</link>
            <guid isPermaLink="false">https://medium.com/p/78aba8413bb6</guid>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[education]]></category>
            <category><![CDATA[cognitive-science]]></category>
            <dc:creator><![CDATA[Lucas Vollet, PhD]]></dc:creator>
            <pubDate>Fri, 29 May 2026 23:06:28 GMT</pubDate>
            <atom:updated>2026-05-29T23:06:28.229Z</atom:updated>
            <content:encoded><![CDATA[<p>Logical form was always an attempt to figure the causality of thought. Not causality in the physical sense, but a particular mechanics, the sustainable movement from condition to consequence. The kind of movement where the conclusion does not merely follow but could not have failed to follow given what came before. Where the connection is not accidental but load-bearing. Where something is genuinely preserved across the transition from premise to conclusion.</p><p>Logical form was historically attractive because it seemed to expose the machinery that made the necessity of this transition visible. The syllogism, the quantifier structure, the proof rule: these were attempts to reveal what was doing the work.</p><p>Carnap’s project in the Logical Syntax was not simply to describe how scientists talk. It was to show that an axiomatic system — a non-bifurcating system, one where valid transformations never split into rival paths of equal standing — could fix the movement from condition to consequence in a machine that works predictably. That meaning specification could be anchored. That the sustainable movement logical form had always been tracking could be captured in finite, well-structured steps.</p><p>For Carnap, especially in the <em>Logical Syntax</em> period, logical form is still partly justificatory. He thinks that once a linguistic framework is fixed, the consequences are objectively determined by the rules of the framework. The framework is not merely describing our inferential habits; it constitutes the standards by which consequences are licensed.</p><p>When the framework efficiently organizes the transition — Quine’s position — the necessity you feel is a very reliable habit. A trained response. The framework is the best available engineering solution for keeping predictions coherent and inferences manageable. It works. But notice the change: it is a model of the inference. Not a normative inevitability. The transition is subtle and that is why many feel like Quine is making an unnecessary fuss.</p><p>The fact is: once you move to efficient organization, the question Quine keeps asking becomes unavoidable: what explanatory work is the extra machinery doing? Intensionality, analyticity, meanings as objects, propositions as entities — all of this was built to explain why necessity felt like more than habit. To show that the transition was not just reliable but licensed. Quine’s challenge is: show me what that addition explains that the engineering solution does not already handle. And if you cannot show me, the extra machinery is not explaining anything. It is just restating the phenomenon in more expensive vocabulary.</p><p>Now lets go further with Quine’s issue and why it is our issue as well. The problem is not merely the cost of Carnap’s ontology of superidentities — such as synonyms and other cross-contextual identities — but the fact that Carnap hardcodes that very machinery as an a priori way of separating sense from nonsense, science from metaphysics, and what can be spoken with feedback from what the system simply does not process. Once he does that, logic becomes an extraordinary, unreviseable fact. Yet any genuine enrichment of the system must be paid for in terms of a reorganization of the prior space of semantic revision; otherwise, we would not be learning new conceptual tools but merely reviving them, like Platonic reminiscences.</p><p>When he fixes the boundary between sense and nonsense at the level of the framework itself, something that looks like clarity begins to function like dogma.</p><p>And then we come to know why Quine’s version of the workings of logic is more faithfull to what learning requires. If the system genuinely learns — if it acquires new conceptual tools rather than merely unpacking what was always already implicit in its initial setup — then that enrichment must be paid for by the reorganization of a prior space of semantic revision. Something in the framework must be moveable. Some boundary that was functioning as constitutive must become revisable in light of what the enriched system can now do.</p><p>Carnap’s hardcoding forecloses that. It makes logic exceptional not because it has earned exceptionality but because the architecture decided in advance where the ground was.</p><p>The deeper difficulty is not simply that Carnap sought to formalize consequence, but that the very machinery responsible for fixing consequence is itself largely insulated from revision. Logical Syntax was intended to provide a principled distinction between sense and nonsense, science and metaphysics, legitimate inference and empty speculation. Yet once these criteria are embedded into the architecture of the framework itself, the framework begins to occupy a privileged position with respect to the very revisionary pressures it governs. The result is that logical form risks becoming an extraordinary and unexplained fact: the one component of the system that organizes semantic traffic without itself being subject to semantic reorganization. This creates a tension with the history of conceptual innovation. Scientific progress rarely consists merely in deriving new consequences from an unchanged inferential apparatus; rather, it often involves the emergence of new conceptual tools that reorganize the very space of possible consequences. Non-Euclidean geometry, relativity, quantum theory, and non-classical logics did not simply enrich an existing framework — they altered the conditions under which inferential work could be performed. If such enrichments are not paid for through a reconfiguration of the prior semantic landscape, then learning becomes difficult to distinguish from recollection.</p><p>This is precisely where attempts to scan brains for the productive workings of cognition fail: because the conceptual challenges that truly matter do not behave like schoolroom rituals, where every step follows a preset formula. Once a new solution to a problem arises, or a proposition reorganizes the debate field, or someone moves the unification of a field forward, none of that can be legitimately challenged from the outside by saying, “Hey, you didn’t follow the formula the way my prior framework teaches.” Whether one finds a solution to the a mathematical problem using AI, or makes other big conceptual intervention to a longstanding debate, questioning that intervention from the position of a prior framework — or from the blind spot of the framework, like the Carnapian insisting, “This I cannot recognize because this language could not predict it” — is a small-minded way of treating learning and intellectual production. And yet, this is precisely the deep structure of virtually all anti-AI anxiety you see right now — from the institutional base to what presents itself as the intellectual ceiling. The MIT study that circulated recently. The academic statements. The faculty memos. The LinkedIn posts. They all share the same architecture. They are frameworks insisting on their own constitutive status in the face of enrichments that reorganize the space those frameworks were built to govern.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=78aba8413bb6" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The Apocalyptic Parrot: On a Contradiction That Knows What It Is Doing]]></title>
            <link>https://medium.com/@lucasvollet/the-apocalyptic-parrot-on-a-contradiction-that-knows-what-it-is-doing-57af5d4e3920?source=rss-e184748bf401------2</link>
            <guid isPermaLink="false">https://medium.com/p/57af5d4e3920</guid>
            <category><![CDATA[philosophy-of-mind]]></category>
            <category><![CDATA[philosophy]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[writing]]></category>
            <dc:creator><![CDATA[Lucas Vollet, PhD]]></dc:creator>
            <pubDate>Mon, 27 Apr 2026 10:45:04 GMT</pubDate>
            <atom:updated>2026-04-27T10:45:04.801Z</atom:updated>
            <content:encoded><![CDATA[<p>In this post I want to tackle again a persistent contradiction. It is not clear how much of it represents antagonistic routes competing for the same ground, or whether there is a way out through some future synthesis. But it is clear that a deep internal confusion is doing the heavy lifting here — carrying conceptual weight in a direction that is deeply inconsistent and yet somehow captures the exact logic of a hysterical problem sustaining its own denial.</p><p>Which is interesting. Because we just came through a sequence of very public, very proud diagnoses of collective irrationality — flat earthism, covid denial, vaccine panic — each one carefully documented, pedagogically unpacked, held up as a lesson. And this one is operating as if none of that happened. As if the educational moment left no residue whatsoever.</p><p>The contradiction appears in various forms. But there is one version of it that becomes irresistible.</p><p>People are grounding their entire reassurance culture — their professional identity, their self-respect, their sense of the future — on the conviction that large language models are trivial parrots doing mechanical work. That the whole AI industry is a bubble built on false deliveries and collective hallucination. Fine. That is a position. It can be argued.</p><p>But those same people are simultaneously running an incessant campaign to inflate the very bubble they claim to see through — endlessly amplifying AI’s powers to cheat, to disrupt, to deceive, to replace, to destroy entire categories of human work and meaning. Which produces a strange creature: a trivial parrot with apocalyptic powers. A mechanical toy that is somehow above human reason when we need to evaluate its outputs, and below it when we need to justify our comfort.</p><p>The machine is stupid enough to dismiss. And powerful enough to blame for everything.</p><p>That is not a position. That is a symptom. And the symptom is interesting precisely because it is doing exactly what the flat earther and the covid denier did before it — using the energy of the threat to fuel the denial of the threat, keeping both alive simultaneously, because resolving the contradiction would cost something neither side is willing to pay.</p><p>And since transparency about my own position might add something to the trust of whoever is reading this far:</p><p>I am fully aware of the power of this machine to disrupt old sociological prestige hierarchies and the models by which epistemic competence has traditionally been measured and rewarded in education. But I am also confident that this is no reason to sacralize the old sociological status quo as if it were a metaphysical default.</p><p>As for the epistemic problem itself — the anxiety about authenticity, authorship, original thought — I think it fades considerably as soon as we move from one register to another. From A: local, bounded, easily pipeable forms of student testing and individual output verification. To B: large conceptual challenges — scientific problems, theoretical breakthroughs, genuine synthesis across disciplines — the kind of work that is much harder to route through an innocent pipeline of personal production or ego-driven necessity of proving who did it.</p><p>The machine is genuinely competitive at A. It is not, I think, at B. Humans will remain above the machine at the level that counts.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=57af5d4e3920" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Notes on a Debate That Already Happened (the AI-debate again)]]></title>
            <link>https://medium.com/@lucasvollet/notes-on-a-debate-that-already-happened-the-ai-debate-again-df30de4f58ca?source=rss-e184748bf401------2</link>
            <guid isPermaLink="false">https://medium.com/p/df30de4f58ca</guid>
            <category><![CDATA[writing]]></category>
            <category><![CDATA[philosophy]]></category>
            <category><![CDATA[ai]]></category>
            <dc:creator><![CDATA[Lucas Vollet, PhD]]></dc:creator>
            <pubDate>Sat, 25 Apr 2026 22:48:54 GMT</pubDate>
            <atom:updated>2026-04-25T22:48:54.913Z</atom:updated>
            <content:encoded><![CDATA[<p>There is an old assumption built into how we judge thought. It’s a convenient myth. But not a useless one.<br>Because it overlaps with something real: we do have an intuitive sense of when something feels like thinking, and when it doesn’t. A thought stops feeling like thinking when its hidden assumptions do all the heavy lifting unnoticed, and we instinctively reach for words like reflex, conditioning, program. Now people are too eagerly adding “machine-assisted” to that same list. But the problem runs deeper than anyone currently arguing about it seems to realize, and the debate they think they’re having is one philosophy settled, unsettled, and resettled over centuries. They’re reinventing the wheel, badly, from scratch. Actually, Philosophers have noticed this for centuries. What is new is not the phenomenon, but its visibility. With AI in the picture, more people are starting to detect the difference, not because machines introduced it, but because they made it harder to ignore. Two things need to be settled before this rediscovery starts looking embarrassing.<br>First: not using AI won’t save you from the “non-thinking” list. If your thought routinely runs on programs it can’t recognize, examine, or name — the instrument is irrelevant. You’re already there.<br>Second: there isn’t even scientific consensus on what the hidden assumptions, categories, or implicit structures of (an ideal) genuine thinking actually are, let alone a measure for when they cross the line into reflex or conditioning. The phenomenological ground for spotting “ideal thinking” in opposition to mere “psychological reflex” here is contested, and it has been contested for a long time. So hold the horses. You didn’t discover a new continent. The maps already exist. They’re just harder to read than a Twitter thread.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=df30de4f58ca" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[AI, Philosophy, and the One Discipline AI Cannot Displace]]></title>
            <link>https://medium.com/@lucasvollet/ai-philosophy-and-the-one-discipline-ai-cannot-displace-e21e7ab0e3e6?source=rss-e184748bf401------2</link>
            <guid isPermaLink="false">https://medium.com/p/e21e7ab0e3e6</guid>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[philosophy]]></category>
            <dc:creator><![CDATA[Lucas Vollet, PhD]]></dc:creator>
            <pubDate>Thu, 23 Apr 2026 21:17:33 GMT</pubDate>
            <atom:updated>2026-04-23T21:17:33.019Z</atom:updated>
            <content:encoded><![CDATA[<h3>I have noticed a strange ironic move in recent reports about AI displacement of jobs. The usual targets are all there — the analyst, the paralegal, the junior coder. Fields with clear economic footprints, clear reasons to feel threatened.</h3><p>But now they also target: philosophy.</p><p>It may be because some people formed in this area got a little too cocky in the last few years — the philosophy grad turned tech ethicist, the confident op-ed explaining what AI “really means.” Perhaps some correction was warranted. Perhaps.</p><p>But the move is still remarkable. Take Alex Karp — CEO of Palantir, holder of a doctorate in philosophy from Goethe University Frankfurt — who recently declared that philosophy degrees are “very hard to market” in the age of AI.</p><p>The problem is that he picked the one target that already made peace with its own economic irrelevance centuries ago<strong>.</strong> Socrates didn’t have a LinkedIn. Philosophers were never inside the economic machine to begin with. The philosophy department was not going to be disrupted by AI because it was never really connected to the circuits AI is running through. What makes it more funny: philosophy loses nothing with AI revolution. And it can actually gain one thing or two. So the line trying to link AI and philosophical obsolescence doesn’t just fail to correlate. It runs in the opposite direction.</p><p>The genuinely uncomfortable targets would have been the data analyst, the mid-level consultant, the junior lawyer — people who are actually Palantir’s customers. Those he left alone.</p><p>What makes it funny is what Karp says <em>survives</em>: neurodivergent thinking. People who escape conventional frameworks, who sit comfortably inside unresolvable problems. That is, more or less, a description of philosophy done seriously. Wittgenstein’s questions about language and meaning are now live engineering problems that nobody in AI knows quite how to solve.</p><p>Karp built a $433 billion company on problems that Wittgenstein already had a vocabulary for.</p><p>And he did it with a philosophy doctorate — which he used to explain why philosophy doctorates are useless. The recursion, apparently, went unnoticed.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=e21e7ab0e3e6" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Quine, AI, and the Myth of Cognitive Erosion]]></title>
            <link>https://medium.com/@lucasvollet/quine-ai-and-the-myth-of-cognitive-erosion-24c749c6a1cf?source=rss-e184748bf401------2</link>
            <guid isPermaLink="false">https://medium.com/p/24c749c6a1cf</guid>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[ai-and-education]]></category>
            <category><![CDATA[education]]></category>
            <category><![CDATA[education-technology]]></category>
            <category><![CDATA[cognition]]></category>
            <dc:creator><![CDATA[Lucas Vollet, PhD]]></dc:creator>
            <pubDate>Sat, 18 Apr 2026 12:24:18 GMT</pubDate>
            <atom:updated>2026-04-18T12:24:59.328Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ReD1e5gEz8OiAR3Y2kLedg.jpeg" /></figure><p>According to W. V. O. Quine, meaning has no inner code and no privileged origin. An identity statement holds not because of hardcoded necessity, but because it has been sedimented into a web of belief:stabilized through use, revision, and feedback across time.<br>This is not just a semantic thesis. It quietly underwrites large parts of contemporary cognitive science: predictive processing, neuroplastic learning, dynamical systems approaches. In all of them, cognition is not the activation of a fixed inner grammar, but an ongoing recalibration, sensitive to error, environment, and history.<br>The discourse of “cognitive erosion” presupposes exactly what this framework denies: a stable inner faculty, a kind of intellectual muscle that can decay when not properly exercised, or when “outsourced” to machines. <br>But in a Quinean picture, there is no such muscle. There is no untouched core to erode. What we call learning is already a process of displacement: old beliefs overwritten, networks reweighted, significance redistributed. The system does not lose purity. It transforms structure.<br>This cuts directly against the conservative framing.<br>AI assistance is often sold as a form of intellectual laziness, a prosthetic for the cognitively weak. But under Quine’s view, this accusation misfires. If cognition is already distributed, revisable, and environmentally scaffolded, then AI is not a degeneration of thought, it is simply another layer in the feedback loop.</p><p>In the world of educational science, I think this view helps frame two competing paradigms.</p><p>One — what I call the “student as possible criminal” model — operates under constant suspicion. It fears losing control if it cannot track, audit, and certify every step of epistemic competence. It needs to keep contributions rigidly separated, because if teacher and student begin to appear as co-participants in the construction of knowledge, the hierarchical distinction becomes harder to justify. Learning, in this model, is something to be policed.</p><p>The other treats knowledge as a collective and distributed construction, something that emerges through interaction, mutual correction, and coupling with tools, rather than through the vertical transmission of an “inner code” supposedly implanted after reading a book. Here, the value is not in isolating the origin of each contribution, but in how the system as a whole gains the capacity to reorganize, respond, and refine itself.<br>Of course, the main “disadvantage” here is that it dissolves any stable sense of hierarchical difference and the myth of authorial prestige.<br>Is that a real loss? Institutionally and sociologically, yes. <br>But not theologically. Not at the level where knowledge was imagined to carry a kind of inner sanctity, a privileged source, or an almost sacred authorship.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=24c749c6a1cf" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>