I have written suspiciously little about music for a blog entitled “Program Notes”. Well, last week — May 7, to be exact — was Johannes Brahms’s birthday (happy 192nd, Johannes!), so it seems a fitting moment to write down some thoughts I have had floating around for a while.
Brahms occupies a curious place in Western musical history. Among the pantheon of “great composers,” he is perhaps the first who was not self-consciously an innovator. (Perhaps Mozart is a counter-example? But if in his youthful period one hears nothing but an almost uncanny perfecting of the Classical style, the mature works of his final decade disclose a latent genius for musical innovation that at times nearly shatters the mold; as with Schubert, one can only imagine how different music would have been had he been given more time.) During his own lifetime Brahms was known as a notable musical conservative, a protégé of the Schumanns and an inheritor of the Beethoven legacy, in contrast to the self-consciously progressive and experimental followers of Liszt (especially Wagner) — the (hyperbolically) so-called “War of the Romantics.” There is no single genre with which he is singularly identified, which he can be said to have (re)invented, transformed, or redefined — unlike Mozart with the opera, Schubert with the song cycle, Liszt with the tone poem, Mahler with the symphony, or Beethoven with pretty much everything (but especially the symphony and the string quartet). Nor is there an instrument whose technique is distinctively and permanently marked by his influence — unlike Bach for the violin and the keyboard, Beethoven and Chopin and Liszt for the modern piano, or Mendelssohn and Wagner (quite differently from one another) for orchestration. His style pays constant homage to Bach’s finely tuned counterpoint, Haydn’s sense of proportion, Beethoven’s ear for the dramatic flair, and Schumann’s expressive melody; but it is hard to say, whatever it might mean to say it, that in any of these areas he “improves” upon his forebears.
Part of all this, no doubt, is that Brahms was a notorious perfectionist — spending nearly twenty years writing and rewriting his First Symphony, and burning the manuscripts of more than a dozen string quartets he considered inadequate. But one can equally say of “perfectionism” that it is an unwillingness to measure oneself by any standards that transcend or relativize those one is given. The First Symphony, after all, took twenty years because it had to be worthy to publish after Beethoven’s Ninth (a burden which Schumann and Mendelssohn had notably not felt). Beethoven’s sheer artistic self-belief (and self-regard) was what permitted him to dispense with the artistic conventions he inherited from Haydn and Mozart, and every great composer after Beethoven considered that to be truly great one must at least try to be like Beethoven in this respect. Every great composer, that is, except for Brahms. He alone seemed to think it worthy to simply and creatively conserve the traditions he inherited, offering to posterity a handful of finely polished gems in which, like the Silmarils of Fëanor, the light of now-past ages is caught and distinctively refracted. And a small handful indeed: in the genres which his great forebears had seen, or had come to be seen, as offering special artistic statements — the string quartet, the piano trio, the piano sonata, and above all the symphony — he left just a few pieces each: string quartets three, piano trios three, piano sonatas three, and symphonies four. If there is a genre in which he was, perhaps, the greatest “innovator” of his day, it is that genre which most self-consciously honors the past: the theme and variations.
All this may sound curiously negative, as though I am suggesting (as Richard Strauss said about himself) that Brahms is “not a first-rate composer, but a first-class second-rate composer.” Not so. Brahms, in his totality, is certainly greater than Strauss (who, as the Brits say, routinely over-eggs the pudding a bit — though that masterpiece of his twilight years, the Vier letzte Lieder, deserves to stand in the first rank). The best passages in Brahms are as transcendently great as anything in Beethoven or Mozart. I am thinking, specifically, of the last five minutes (102 bars) in the first movement of the First Piano Concerto, though there is any number of passages I could spotlight. This movement, and this passage in particular, exemplifies all the best qualities of his writing: the organic expansion of just one or two simple musical cells into a vast whole; a remarkable economy of both counterpoint (there are rarely more than two separate lines moving simultaneously) and orchestration (somehow creating a full, sustained sound without Wagnerian orchestral busy-work); judicious exploitation of the flexible, propulsive rhythms available in his long triple meter, keeping the energy flowing through long yet elegantly balanced melodic lines; and the perfectly seamless, almost invisibly prepared transition from the calm light of the second theme to the darkness of the coda (at bar 438, 22:03 in the above linked recording), like a great cloud slowly obscuring the face of the Sun. There is nothing pretentious, nothing self-serious, nothing indulgent in Brahms. Everything is heartfelt, often even passionate, but utterly sincere. Where Mozart’s music sounds effortless, almost too perfect to be real, and Beethoven’s music sounds immensely effortful, every note as if written with blood — well, Brahms’s music sounds, simply, human: the music of human life, life as really lived, not as larger than life.
Now indulge me as I offer some speculation. In his masterpiece of criticism Real Presences (1989), George Steiner draws attention to the “broken contract” between logos and kosmos, immanent language and transcendent reality: the gulf (so he argues) at the heart of modern humanity’s sense of alienation. If there is no God, there is no “real presence” in anything we say: our words are meaningless. (Steiner himself was, notably, unable to believe in God: throughout Real Presences he writes of God and the transcendent as one who longs for but cannot himself have them.) And it was in the 1870s, Steiner suggests, that European critics and scholars began to advocate for detaching logos and kosmos. I am not even one-thousandth the expert on European arts and letters that Steiner was, but I cannot help noting that in the realm of music, this is precisely the period when tonality and tradition — the so-called “Common Practice” — begins to break down. Wagner’s “Tristan chord” (premiered 1865) is often seen as the touchstone for this development: the first public statement by a leading composer in a major work that the boundaries of tonality and the “rules” of voice-leading could be breached for the sake of expression. Of course, one is not terribly hard pressed to find Tristan-chord-like harmonies and resolutions in earlier composers (Schumann!), but it is hard to deny that there is something… flagrant? iconoclastic? Promethean? in the use Wagner there makes of it. And in any case, the floodgates opened in the 1870s and onward — with Verdi, Franck, Saint-Saëns, and Mussorgsky (all born before 1850) following Wagner in preparing the ground for really major innovations by Puccini, Mahler, Debussy, and Strauss (born after 1850) within a just-barely-tonal paradigm. By the beginning of the 20th century, Ravel and Scriabin were conjuring essentially non-tonal landscapes, and Schoenberg was developing the twelve-tone paradigm by which he sought to banish the concept of a single tonal center from his music — a deliberate repudiation of kosmos in favor of (a highly mathematical and schematized notion of) logos. All that was (apparently) solid melted, in the course of a few decades, into air.
And it is in precisely those decades, in the midst of so much musical chaos, that we find Brahms at work. He is a son of his age, not of some other age; he is not, and cannot be, a mere repristinator. But he is that son not as an innovator, but as a creative conserver, aware of how rich is his inheritance and seeking to make good use of it. What we hear in Brahms is always something new, but never something novel. Perhaps this is the way — the only way? — to flourish in modernity.
Life-works left resonantly unfinished at their creators’ deaths: St. Thomas’ Summa Theologiae; Karl Barth’s Kirchliche Dogmatik; W. A. Mozart’s Requiem; J. S. Bach’s Die Kunst der Fuge.
Our dogmatic labours can and should be guided by results which are venerable because they are attained in the common knowledge of the Church at a specific time. Such results may be seen in the dogmas enshrined in the creeds. But at no place should these replace our [16] dogmatic labours in virtue of their authority. Nor can it ever be the real concern of dogmatics merely to assemble, repeat and define the teaching of the Bible… Exegetical theology investigates biblical teaching as the basis of our talks about God. Dogmatics, too, must constantly keep it in view. But only in God and not for us is the true basis of Christian utterance identical with its true content. Hence dogmatics as such does not ask what the apostles and prophets said but what we must say on the basis of the apostles and prophets. This task is not taken from us because it is first necessary that we should know the biblical basis.
— Karl Barth, CD I/1, 15–16. I have often seen the bolded sentence criticized as reinscribing the “broad and ugly ditch” between Scripture and theology (“Barth thinks you shouldn’t just accept the theology taught by the Biblical writers!”). But it cannot be decontextualized in that sense. There is a distinction operative for Barth, which he has already laid out, between “exegetical theology,” or biblical theology, and dogmatic theology (as well as practical theology: see pp. 4–5). Biblical theology does precisely ask what the Biblical writers taught. Dogmatic theology, in contrast, is the self-critical reflection by the Church on what she teaches: a task which for Barth requires continual return to and testing by the biblical writers, which takes seriously the fact that there is historical difference and development since the time of the apostles and prophets (not all of it benign), and which is never — can, this side of the eschaton, never be — complete.
Lee explained to me that by showing the world AI could be used to cheat during a remote job interview, he had pushed the tech industry to evolve the same way AI was forcing higher education to evolve. “Every technological innovation has caused humanity to sit back and think about what work is actually useful,” he said. “There might have been people complaining about machinery replacing blacksmiths in, like, the 1600s or 1800s, but now it’s just accepted that it’s useless to learn how to blacksmith.”
If only this 19-year-old Columbia suspendee had, at a minimum, done what he apparently did for every assignment and asked ChatGPT for information: “When were machines developed that could assist in metalworking, and have they made the crafts of smithing and metalwork obsolete?” But even asking that question — writing that prompt — would have required a measure of historical literacy, nay, a sliver of interest in history at all.
This (now former) student is an especially egregious offender, worthy indeed of becoming the framing device in a breathless New York Magazine story, but there is nothing remarkable about what he represents: it is the characteristic disease of “move fast and break things” culture. All that is prized is “innovation,” because innovation makes money fast and lets the innovator get out before he (and it is usually a he) is held accountable to clean up the wreckage. The destroyers do not understand, and do not want to understand, the things they are out to destroy.
When one prays repeatedly through the Psalter in sequence, one tends to start noticing patterns in how the Psalms are arranged, or at least suspecting characteristic editorial strategies. The set of Psalms for Evening Prayer on the 28th of the month — 136, 137, 138 — exemplifies one of those strategies: what I think of as the containment of imprecation.
Psalm 137 is perhaps among the most famous of Psalms: “By the waters of Babylon, we sat down and wept, when we remembered Zion,” it opens. We are in the immediate aftermath of the exile from Judah, with the grief still raw, the horror still fresh. “On the willows there we hung up our lyres. For there our captors required of us songs, and our tormentors mirth, saying, ‘Sing us one of the songs of Zion!’ How shall we sing the Lord’s song in a foreign land?” Then the Psalm goes on in progressively darker tones, with (interestingly) two sets of imprecations. First is the less famous double self-imprecation, enjoining the singer not to forget Zion — “If I forget you, O Jerusalem, let my right hand forget its skill; let my tongue stick to the roof of my mouth, if I do not set Jerusalem above my highest joy!” Only then follows a double imprecation against Judah’s enemies. The singers urge the Lord to remember Edom’s complicity in the “day of Jerusalem, how they said ‘Lay it bare, lay it bare, down to its foundations!”, before concluding with the most notorious passage: “O daughter of Babylon, doomed to be destroyed, blessed shall he be who repays you with what you have done to us! Blessed shall he be who takes your little ones and dashes them against the rock!”
There is a lot going on here, even just within the Psalm itself. For starters, the final imprecation — certainly as violent as anything in the entire Bible — is, strictly speaking, not a curse, but a blessing. It performs a complexly layered speech act: 1) pronouncing a (future and hypothetical) blessing on conquering soldiers who will dash Babylonian infants against the rock, which in turn 2a) reveals to the reader, and 2b) reminds the singer, just what the Babylonian soldiers themselves did when they conquered Jerusalem; thus 3) entreating the Lord to mete out his retributive justice (as the previous clause makes clear: “who repays you with what you have done to us”), and only after and through those layers 4) wishing for the violent destruction of the singers’ enemies. This violent pitch is also only reached after the singers have recalled their mockery by their captors (“Sing us one of the songs of Zion!”), and their betrayal by Brother Edom (“Lay it bare!”). None of this makes, or should make, the final sentence itself “easier” to read (or pray), but it does (in a certain literary sense) contain the scope and import of the curse, and illuminates the deep emotional complexity and psychological honesty of the Psalter.
Now observe how the editors of the Psalter contextualize and contain this imprecatory outburst by placing it between Psalms 136 and 138. Psalm 136, first: this is surely among the most uncomplicatedly celebratory Psalms, with its recounting of the Lord’s “great wonders” punctuated by the response “His steadfast love endures forever.” These “great wonders” are, first, the orderly creation of heavens, earth, and waters, and the lights that rule over them; second, the deliverance of the elect nation in the Exodus, their protection through the wilderness (“To him who struck down great kings…”), and their conveyance into the promised land. (There is just the merest hint of Judges-style post-conquest troubles: “It is he who remembered us in our low estate… and rescued us from our foes”.) There is no explicit mention of Jerusalem, but the narrative setting is — at least imaginatively — before exile. And if that is implicit in Psalm 136, it is made explicit in the tightly linked Psalm 135, which concludes in Jerusalem herself: “Blessed be the Lord from Zion, he who dwells in Jerusalem! Praise the Lord!” For that matter, the whole sequence Psalms 120–134 are the “Psalms of Ascent,” sung by pilgrims on their way to the festivals in Jerusalem, renarrating the long journey from “the tents of Kedar” (120:5) to “the house of the Lord” (134:1). The Psalms which precede 137 are, quite literally, “the songs of Zion” demanded by the Judahites’ Babylonian captors.
Meanwhile, in Psalm 138, we are in a tonally somewhat different world than 136 (as well as temporally different, per the superscription “of David”). It is undoubtedly a Psalm of thanksgiving, but the exuberance is tempered by recent suffering and deliverance: “On the day I called, you answered me… Though I walk in the midst of trouble, you preserve my life; you stretch out your hand against the wrath of my enemies, and your right hand delivers me.” Nevertheless, there is at least one indubitable link to 136: that Psalm’s refrain appears again in David’s concluding lines, “The Lord will fulfill his purpose for me; your steadfast love, O Lord, endures forever” — before the somewhat open-ended “Do not forsake the work of your hands.” (The refrain is thus modulated, perhaps, into an injunction that the Lord should be mindful of his own nature and remain faithful to his covenant!) Less overt, but still clear, links to 136 include “all the kings of the earth” giving thanks to the Lord (recalling, by contrast, the kings who were struck down); “though the Lord is high, he regards the lowly” (recalling the Lord remembering Israel “in our low estate”); and, perhaps more speculatively, “before the gods I sing your praise” (recalling the “great lights” created to rule over the day and the night). And our note that (at least) 134–136 are among “the songs of Zion” draws attention to the liturgical setting of 138: “I bow down toward your holy temple and give thanks…”
So what were the editors of the Psalter thinking in placing 137, the paradigmatic Psalm of grief and rage at Jerusalem’s destruction, between 136 and 138, two Psalms of rejoicing in Jerusalem? I suggest that it is purposeful, and strikingly psychologically insightful. They have put grief and rage — and yes, imprecation — in its proper place. The exiles’ anger is given its full venting, as it must be. The sheer horror of violence against the innocent, compounded in the destroying victors’ demonic mockery, must be recalled, and these must continually shock to the point of outrage. There is to be no naïvety. Evil must have its due, and — when revealed for what it truly is — what it is due truly is cursing. But precisely as — and because — the curse is offered up to God, it is given over to God. It is made His responsibility (“Do not forsake the work of your hands”). And as it is made His responsibility, it is contained. It is put into the context of God’s creative and saving blessing. In fact, it only acquires its force from the fact of His creative and saving blessing. And by that same fact the curse is given its definite limitations, limitations which are notably not placed upon the blessing. The imprecation is not allowed to devour the Psalmist from the inside, but it is released; better still, the Psalmist is released from it, to the joy of God’s abundant blessing. Sin always crouches at the door of imprecation, but in offering the imprecation as prayer, the Psalmist masters it.
So the curse is put in its place, and thus we are promised that the curse will not reign forever, but that blessing, having given way to curse, will one day be restored. This yields the larger narrative purpose in the Psalter, of which this psychological purpose is an icon. The sequence of Psalms 136–137–138 enacts in small the story of exile and return. As we read, we follow the Psalmists through time, from exuberant rejoicing through gutting anguish to renewed joy; we experience the dialectic of resonance and alienation; we know the presence of God even as we momentarily feel His absence; we trace what my teacher Jeremy Begbie describes as the movement of “home, away, and home again” — with the essential acknowledgement that home, when you do return, is never quite the same. We rehearse the pattern of the Lord’s faithfulness and steadfast love.
Or, in the words of that great Master Teacher of Scripture: “Was it not necessary that the Messiah should first suffer all these things, and then enter into his glory?”
St. Irenaeus analogizes the diversity-in-order of creation to a musical instrument being played:
[Since] created things are various and numerous, they are indeed well fitted and adapted to the whole creation; yet, when viewed individually, are mutually opposite and inharmonious, just as the sound of the lyre, which consists of many and opposite notes, gives rise to one unbroken melody, through means of the interval which separates each one from the others. The lover of truth therefore ought not to be deceived by the interval between each note, nor should he imagine that one was due to one artist and author, and another to another, nor that one person fitted the treble, another the bass, and yet another the tenor strings; but he should hold that one and the same person [formed the whole], so as to prove the judgment, goodness, and skill exhibited in the whole work and [specimen of] wisdom. Those, too, who listen to the melody, ought to praise and extol the artist, to admire the tension of some notes, to attend to the softness of others, to catch the sound of others between both these extremes, and to consider the special character of others, so as to inquire at what each one aims, and what is the cause of their variety, never failing to apply our rule: neither giving up the [one] artist, nor casting off faith in the one God who formed all things, nor blaspheming our Creator.
— Against Heresies 2.25.2 (ANF, Vol. 1). The principal point here is, of course, the oneness of God (contra the Gnostics). Nevertheless this is a rich analogy, which deserves fuller exegesis.
I am the man who has seen affliction under the rod of his wrath;
he has driven and brought me into darkness without any light;
surely against me he turns his hand again and again the whole day long.
He has made my flesh and my skin waste away; he has broken my bones;
he has besieged and enveloped me with bitterness and tribulation;
he has made me dwell in darkness like the dead of long ago.
He has walled me about so that I cannot escape; he has made my chains heavy;
though I call and cry for help, he shuts out my prayer;
he has blocked my ways with blocks of stones; he has made my paths crooked.
He is a bear lying in wait for me, a lion in hiding;
he turned aside my steps and tore me to pieces; he has made me desolate;
he bent his bow and set me as a target for his arrow.
He drove into my kidneys the arrows of his quiver;
I have become the laughingstock of all peoples, the object of their taunts all day long.
He has filled me with bitterness; he has sated me with wormwood.
He has made my teeth grind on gravel, and made me cower in ashes;
my soul is bereft of peace; I have forgotten what happiness is;
so I say, "My endurance has perished; so has my hope from the LORD."
Remember my affliction and my wanderings, the wormwood and the gall!
My soul continually remembers it and is bowed down within me.
It is only — only! — after these words that the prophet utters the most famous words of this chapter:
But this I call to mind, and therefore I have hope:
The steadfast love of the LORD never ceases; his mercies never come to an end;
they are new every morning; great is your faithfulness.
"The LORD is my portion," says my soul, "therefore I will hope in him."
The LORD is good to those who wait for him, to the soul who seeks him.
It is good that one should wait quietly for the salvation of the LORD.
It is good for a man that he bear the yoke in his youth.
Let him sit alone in silence when it is laid on him;
let him put his mouth in the dust — there may yet be hope;
let him give his cheek to the one who strikes, and let him be filled with insults.
Karl Barth (CD IV.1, §59.1):
In being gracious to man in Jesus Christ, God acknowledges man; He accepts responsibility for his being and nature. He remains Himself. He does not cease to be God. But He does not hold aloof. In being gracious to man in Jesus Christ, He also goes into the far country, into the evil society of this being which is not God and against God. He does not shrink from him. He does not pass him by as did the priest and the Levite the man who had fallen among thieves. He does not leave him to his own devices. He makes his situation His own. He does not forfeit anything by doing this. In being neighbour to man, in order to deal with him and act towards him as such, He does not need to fear for His Godhead. On the contrary… God shows Himself to be the great and true God in the fact that He can and will let His grace bear this cost, that He is capable and willing and ready for this condescension, this act of extravagance, this far journey. What marks out God above all false gods is that they are not capable and ready for this. In their otherworldliness and supernaturalness and otherness, etc., the gods are a reflection of the human pride which will not unbend, which will not stoop to that which is beneath it. God is not proud. In His high majesty He is humble. It is in this high humility that He speaks and acts as the God who reconciles the world to Himself.
Johann Heermann (tr. Robert Bridges):
Who was the guilty? Who brought this upon Thee?
Alas, my treason, Jesus, hath undone Thee.
’Twas I, Lord, Jesus, I it was denied Thee!
I crucified Thee.
Lo, the Good Shepherd for the sheep is offered;
The slave hath sinned, and the Son hath suffered;
For man’s atonement, while he nothing heedeth,
God intercedeth.
For me, kind Jesus, was Thy incarnation,
Thy mortal sorrow, and Thy life’s oblation;
Thy death of anguish and Thy bitter passion,
For my salvation.
Note: Hand over heart, I substantially drafted this post — including its core conceit — long before reading Josh Brake’s latest Substack post. No plagiarism here!
I have two sorts of problems with “AI” in general and Large Language Models (LLMs) in particular. One is the (infinitely ramifying) ethical problem. LLMs are built on deception. They are not human (and not “alive”), do not possess human cognitive faculties, and cannot “know” anything in the ordinary human sense of that word, and yet their model is built on — after vacuuming up an enormous amount of human-created linguistic “content” — mimicking human cognition and knowledge to such an effective degree that you spend all your time relying on GPT-4o or what have you, rather than other human beings. I take this to be a fairly straightforward form of deception, and because of the incommensurability of truth and falsehood, this first problem to be the most fundamental. What does constantly being deceived, and constantly self-deceiving, do to a human being? In what ways are we damaging, and might further damage, ourselves by using such a false tool? (See also: Mammon.) But that’s for another post.
The principles of scientific forestry [TC: planting a single “crop,” in evenly-spaced rectangular grids, in place of the old ecologically diverse forests] were applied as rigorously as was practicable to most large German forests throughout much of the nineteenth century. The Norway spruce… became the bread-and-butter tree of commercial forestry. Originally [it] was seen as a restoration crop that might revive overexploited mixed forests, but the commercial profits from the first rotation were so stunning that there was little effort to return to mixed forests… Diverse old-growth forests, about three-fourths of which were broadleaf (deciduous) species, were replaced by largely coniferous forests in which Norway spruce or Scotch pine were the dominant or often only species. In the short run, this experiment in the radical simplification of the forest to a single commodity was a resounding success… the negative biological and ultimately commercial consequences of the stripped-down forest became painfully obvious only after the second rotation of conifers had been planted… An exceptionally complex process involving soil building, nutrient uptake, and symbiotic relations among fungi, insects, mammals, and flora—which were, and still are, not entirely understood—was apparently disrupted, with serious consequences. Most of these consequences can be traced to the radical simplicity of the scientific forest. … Apparently the first rotation of Norway spruce had grown exceptionally well in large part because it was living off (or mining) the long-accumulated soil capital of the diverse old-growth forest that it had replaced. Once that capital was depleted, the steep decline in growth rates began.
To apply the analogy: Maybe, just maybe, you can implement LLMs without too many problems in the first generation, among a population of adults who have already been educated. Their values have already been formed; they have already learned to read and write and think critically. (This already concedes far too much to the “AI” boosters, but for the sake of the argument, we will not pause overlong.) Perhaps they really could achieve the stunning productivity growth which we are constantly promised (though so far the results don’t seem great!). But even if that were true, can you expect those gains in the second generation, among children who are still being educated? Or would you rather expect systemic failure to ever form values, to learn critical thinking, essential reading comprehension, and basic writing skills? The adults who received pre-LLM educations have an existing store of cognitive and intellectual capital on which to draw as they encounter and learn to use LLMs. But children who never experience education without LLMs will never have the chance to develop that capital.
Furthermore, the broader environment in which this “first rotation” is encountering LLMs is not remotely the same as that in which the “second rotation” will encounter them. Indeed, the environments are being treated as if they are the same, when they should be different. My local school district is now integrating “AI” into primary and secondary education, because “universities and employers will expect AI literacy” — what tool is easier to learn to use than a natural language chatbot? Now, the workplace may appropriately demand certain kinds of efficiency from adult workers, and LLMs may just prove their usefulness in such cases (though in my view the jury is still out). Education, by contrast, should be inefficient, frictional, resistive. The mind is like a muscle: in order to grow, it must be repeatedly stretched to the limits of its capacity. The LLM chatbot is the ultimate anti-friction, super-efficient (except in, you know, water and energy) machine, which promises that you will never encounter resistance ever again; with the new “reasoning” modules, you’ll never have to think for yourself again. The implications for education hardly need to be spelled out.
Scott continues:
As pioneers in scientific forestry, the Germans also became pioneers in recognizing and attempting to remedy many of its undesirable consequences. To this end, they invented the science of what they called “forest hygiene.” In place of hollow trees that had been home to woodpeckers, owls, and other tree-nesting birds, the foresters provided specially designed boxes. Ant colonies were artificially raised and implanted in the forest, their nests tended by local schoolchildren. Several species of spiders, which had disappeared from the monocropped forest, were reintroduced. What is striking about these endeavors is that they are attempts to work around an impoverished habitat still planted with a single species of conifers for production purposes. In this case, “restoration forestry” attempted with mixed results to create a virtual ecology, while denying its chief sustaining condition: diversity.
I leave the resonances between this virtualized ecology and the state of education today as a trivial exercise for the reader.
(Scott’s remarks here of course have many parallels. Ivan Illich makes a remarkably analogous argument, with respect to medicine, in the opening of Tools for Conviviality; and Michael Polanyi offers a structurally similar observation about the Enlightenment “critical movement” that sought to banish belief from knowledge: “its incandescence had fed on the combustion of the Christian heritage in the oxygen of Greek rationalism, and when this fuel was exhausted the critical framework itself burnt away.”)
“Artificial general intelligence,” defined as “a computer able to do any cognitive task a human can do” — as envisioned for example in this new work of science fiction — is computationally impossible to achieve.
This is because “intelligence” — in the sense of “normal human intelligence,” which is presupposed by the above definition of “AGI” — is a) impossible to fully and simultaneously articulate (hereon inarticulable) and b) non-deterministic, and therefore in at least two senses strictly non-computable.
The inarticulability of intelligence has (at the very least) to do with its embodied and relational aspects. “Mind” is neither identical with nor even co-extensive with “brain activity”; rather, “mind” is (to crib from Dan Siegel’s definition) is an embodied and relational process. Emotion in particular seems, as far as the causality can be determined, to be body-first, brain-second, such that it is only articulable after the fact (and in a way that changes the emotional experience). Michael Polanyi’s great work demonstrates in a philosophical register what musicians, artists, and craftspeople have always known intuitively: that the “cognitive task” of playing an instrument or using a tool depends on integrating the instrument or tool into one’s bodily experience, in an inarticulable way. And relationship through interaction with other embodied minds is such a complex process, with so many emergent layers, that not only is it poorly theorized or modeled now, it may be impossible to exhaustively theorize or model — especially because it primarily seems to take place in and through the pre- and in-articulate dimensions of cognition.
Meanwhile, the non-determinism of intelligence has (at the very least) to do with quantum randomness effects in the brain, which at the mesoscale (the level at which daily human, and most complex organic, life takes place) emerge into relatively well-understood and predictable patterns, but at the nanoscale (the relevant level for a hypothetical deterministic model of cognition) are by definition impossible to predict, or even observe without altering them. I am unaware of any good reason to think the quantum effects in, say, an extremely large and inorganic GPU farm, would be interchangeable with or even meaningfully similar to those in a three-pound organic human neural system.
What is computationally possible, as far as I can tell, is a (relatively) high-fidelity simulation of one aspect of human cognition: the comparatively deterministic, hyper-articulated aspect of human cognition which Iain McGilchrist identifies as characteristic of the left hemisphere (hereon LH) of our brains (subject, of course, to obvious caveats from theses 2–4). Note: I am not saying, and I do not take McGilchrist to be saying, that a fully-computed model of the LH itself is possible; only that its characteristic thought-style can be simulated in high fidelity, precisely because that thought-style is comparatively deterministic and hyper-articulated.
In currently existing frontier Large Language Models (LLMs), I take it something like this has already been achieved. Commercially available LLMs are now (to use a technical term) pretty good at processing and reproducing both written and spoken natural language — albeit in such a sterile “voice” that it renders the phrase “natural language” almost meaningless — and quite good at analytically processing huge quantities of formally similar information. These are two of the characteristic specializations of LH cognition, and I expect the next generation of LLMs to be significantly better on both fronts. Notably, some of the persistent failure modes of LH cognition and of LLMs are startlingly similar: “hallucination” or fabrication of nonexistent supporting evidence, a predilection for lying or circumventing rules in order to achieve a desired result, an inability to attend to wholes at the expense of parts, and so forth.
Because much of contemporary Western life (as McGilchrist and others have extensively documented) is already organized to systematically advantage that aspect of human cognition, it is therefore no surprise or, in a sense, any remarkable accomplishment that frontier models now perform at the level of PhD students in solving advanced physics problems (albeit ones with solutions known to currently existing physics), or that some chatbots now “pass the Turing Test.” This is the natural end result of reimagining science as “knowledge production” and credentialing scientists accordingly, or of technologically reducing the typical person’s normal experiences of and capacity for conversation to so great an extent that we now take what the LLMs offer to be “human” conversation. This — and all the attendant social/economic disruption (about which more below) — is all possible without “AGI” itself being computationally feasible.
The second strike against the possibility of “AGI” comes from limits in physical resources. Achievements in LLM development up to this point have been enabled by energy use, water depletion, and resource extraction on an already massive scale. The anticipated investments required for “AGI” (e.g., according to AI 2027, $2 quadrillion in new data centers over the next 10 years!!!) will require exponentially more energy, water, and mineral resources that we either simply do not have on this planet or cannot physically extract from it at the desired rate (unless we invent, say, cold fusion). This is to say nothing of the land required to build all of the new infrastructure. I therefore anticipate that “AI” development will, as a function of resource scarcity, fail to get anywhere close to the scale of investment theoretically required for “AGI.” This may only become clear to “AI” developers, however, after they have already inflicted genuinely ruinous and probably irreversible damage to the environment and to the communities that depend on it.
Considering all this, I find it probable that without ever achieving “artificial general intelligence” as imagined in science fiction, advances in “AI” over the next several years will make all but the top 1–5% of current “symbolic capitalists” functionally obsolete. This includes both high-status sectors such as consulting, finance, advertising, software development, law and legal services, etc., and lower-status (or at least lower-paying) sectors such as journalism, copywriting, teaching, administration, graphic design, the social sciences, etc. (Note that several of these lower-status professions are ones which the Internet revolution has already been destroying.) By “functionally obsolete” I mean that it will almost always be more cost-effective, and nearly as useful, to “employ” an “AI agent” for a task that previously required one to hire a human being.
Sectors that are symbolic-capitalism-adjacent but require long training in embodied skill — e.g., healthcare, the experimental sciences, mechanical engineering, war — will not be functionally obsoleted, at least not so thoroughly. An inorganic robot will never be able to perform skilled tasks in the real world with the same level of ability as a trained human being (see (3) above)… and “organic robots” capable of such skill would pretty much just be, well, lab-grown humans, with many of the same inefficiencies and time-delays as regular humans. (Only a conspiracy theorist would see current Silicon Valley investments in IVF, genetic selection and editing, and artificial wombs as an attempt to create the conditions of possibility for lab-grown humans… right???) But some current features of jobs in these sectors — the features, that is, which are most akin to “AI” core competencies — will be permanently outsourced to “AI agents.”
The “trades” and the “crafts,” on the other hand, will not become thoroughly automated, though they will be in various ways automation-directed and -augmented. Machine maintenance and repair, for instance: machine failure might be AI-diagnosable, but the intuitive skill necessary for actual repairs will remain the province of humans. To deal with water, you’ll always need a plumber. Reality has a surprising amount of detail, and fields like construction and mining will always require meaningful and skilled human attention to reckon with that detail. Agriculture represents an interesting test case: a field that is currently extremely mechanized, but as the lowest-skilled tier of human labor becomes (out of necessity) far cheaper to “buy,” one which may reabsorb much of that excess labor capacity. At the more humanistic end of the spectrum, traditional crafts might make a comeback of sorts (similar to the vinyl resurgence), and the performing arts will always be the province of human beings, though probably far fewer people will be performing artists in fifteen years than are right now; in both cases patronage will be the only economically viable model. For the ultra-wealthy, owning or sponsoring something evidently made only by humans will be a status symbol.
In sum: I believe we are headed neither for the existential-risk, civilization-ending disaster scenarios envisioned by the “AI Doomers,” nor for the golden era of peace and prosperity and universal basic income envisioned by the “AI optimists.” (Where, exactly, do the optimists think the value creation for UBI will come from in an era of mass human unemployment?) Rather, I suspect in the near-ish term we are headed for a poorer, less culturally vibrant, less highly educated world with much greater wealth inequality. This will be a world in which many more people, including some who might otherwise have been symbolic capitalists, work in various kinds of manual labor or “trades”: agriculture, mining, energy, construction, maintenance. Others will depend, one way or another, on the patronage of the few ultra-wealthy. The whole service-economy apparatus that depends on a large leisure class will be semi-permanently diminished in proportion. It might, in other words, look in certain ways remarkably like the period of transition into the Industrial Revolution.
Over the long run, I believe in the resilience of humanity, chiefly because I believe in the faithfulness of God to His wayward creatures. We will not be destroyed or superseded by a higher form of intelligence, nor will we manage to completely destroy ourselves. We are remarkably adaptable and creative: life always finds a way. But we will find that the remarkably widespread prosperity of the last few decades in particular and the last two centuries in general is not, once unlocked, a permanent and automatic feature of human existence. It has depended on our irretrievably consuming the planet’s resources at an ever-accelerating rate. What cannot go on indefinitely must eventually stop. The mechanization snake will finally eat its own tail. The only question is how soon.
Addendum (08.15.2025): Well, this has had much more of an afterlife than I expected. And we have had four months more of AI development (i.e., several lifetimes). What do I think now about all of the above? So far, I feel quite as confident about items #1–7 as the day I wrote them. I am nearly as confident about #8, though I recognize that energy/water use per query is something of a moving target and that many of the major AI developers are investing in energy solutions. (Here’s a science fiction scenario for you: What if we came up with, say, a non-destructive way to siphon geothermal energy from the Yellowstone supervolcano, powering everyone’s Claude Code instances for eons while diverting a real civilizational X-risk?) My greater environmental concern remains the resource extraction — and corresponding ecosystem degradation/destruction, at points of both origin and destination — necessary to build the proliferating data centers that will be necessary to build more and more capable models, and to supply the infrastructure upgrades that get water and energy to those data centers. (This is to say nothing of the environmental justice questions about where data centers are actually built.) The typical American utility depends on infrastructure that is decades old, whose construction was predicated on a long-deceased funding model (i.e., massive federal subsidies), and which is now desperately in need of repair at practically every point. It’s like the old joke about True Libertarians: How will they drive wherever they want, as fast as they want, if there aren’t any roads? Color me skeptical that AI developers will actually prioritize massive upgrades to local infrastructure in places where they build their data centers, though I would love to hear stories to the contrary.
The big questions come, of course, with #9, and perhaps to a lesser degree #10–12. I have looked back on #9 a number of times and thought: “What was I thinking?” 95–99% of current symbolic capitalists is a pretty large number! So let me offer a couple of qualifications and clarifications. First, by saying “probable,” I think I meant “more than 50% probability.” (In my lexicon, I think I use “unlikely” for <10%, “possible” for 25–50%, “likely” for, oh, I don’t know, 60–75%, and “almost certain” for >90%.) I would have been comfortable saying then, and maybe still am, that it is “almost certain” 50% of current symbolic capitalists will be functionally obsoleted (more on that term in a second). Note that this is not the same thing as Dario Amodei’s “half of all entry-level white-collar jobs.” There are a lot of even entry-level white-collar workers — especially in healthcare — who are not pure symbolic capitalists, in Musa al-Gharbi’s sense, even if they presently do a great number of symbolic-capitalist things in the course of their everyday work; my #10 was meant in part to correct against this misconception. And, even in the most successful AI development scenario I can reasonably imagine, functional obsolescence of the bottom 95% of symbolic capitalists does not mean that all those people go “oh well, guess I’d better become a farmhand.” It does mean they move down the prestige and remuneration ladder in various ways, but probably does not mean they have to leap off and start again from the bottom. It also means that fewer young people even aspire to enter those fields in the first place, instead proactively entering less automatable fields (anecdotally, at all colleges outside the absolute most elite tier, student interest in nursing programs is just skyrocketing). And it means that the gates behind which the true symbolic-capitalist jobs are kept become ever more difficult to unlock. A PhD has long offered no guarantee of success, or even of a job, in any field. Now some of the most ostensibly prestigious PhDs are becoming, literally, impossible to get.
“Functional obsolescence” is also an important term for my argument in thesis #9, one which I did not define in quite the way I should. My definition was that “it will almost always be more cost-effective, and nearly as useful, to ’employ’ an ‘AI agent’ for a task that previously required one to hire a human being.” The term “cost-effective” should be clear. “Useful” is obviously less so. What I basically meant is that, to a manager, the short-term quality of work that an “AI agent” can do for the entry-level tasks that require little contextual knowledge will seem so impressive that they will, by default, choose to defer or forgo the long-term gains of adding another human person to the team. In any organization, a new hire, no matter how intelligent or experienced, needs some time to learn the ropes, and depending on the context it may be not just weeks or months but years before the institution sees its return on investment, as it were. (There is a different calculus for primarily internal-facing and primarily external-facing roles here; I suspect the functional obsolescence of the sales account manager, whose personal touch with the client is in fact part of the package sold to the client, will be far slower to arrive than the functional obsolescence of the sales account manager’s executive assistant.) This does not prevent a company having an explicit value on hiring and developing human employees, such that they are willing to forgo the short-term advantages of AI agents. I am immensely grateful to work for one such company. I suspect there are many other symbolic-capitalist companies that are still talking this people-centric talk, but are quietly making preparations to cease walking the walk as soon as it becomes practical to take a different path.
The other main reason that #9 seems far less plausible to a lot of people now than it perhaps did four months ago is the stubborn persistence of certain kinds of AI failures and hallucinations. It still makes up references to nonexistent sources (though less often than it used to); it still offers impossible and logically laughable ex post facto rationalizations for “conclusions” at which it claims to have “arrived” via “chains of thought,” thus obscuring its inner workings (though, admittedly, real human beings are awfully good at this as well); it still can’t reliably tell you how many B’s there are in the word “blueberry” or how many U.S. state names contain the letter R. The recent much-hyped launch, and notable failures, of OpenAI’s GPT-5 has Maria Sukhareva announcing that “AI Winter is coming.” And indeed, if this indicates that we are reaching the upper bound of the logistic curve and the rate of progress is leveling off, then “AI agents” will not be anywhere as good “in the next several years” as is necessary to functionally obsolete 95–99% of current symbolic capitalists. They will still be an amazing technological feat by any metric, but like many such feats of the industrial era, they will not actually decrease the absolute amount of work there is to be done, but simply reshape much of the existing work around themselves. They will fail to be transformative in the way that, say, the telegraph was to the nineteenth-century general or journalist (completely reshaping the possibility structure in which their work is done), instead becoming something more like the washing machine to the mid-century housewife (removes a source of drudgery, freeing up time for… other kinds of drudgery). I was deliberately vague in saying this could happen in “the next several years” — recall: the transformer architecture on which our contemporary LLMs are built was invented in 2017, which in my lexicon is only “several years ago”! — so let’s wait until, say, 2032 to render final judgment. Unfortunately, my intuitions here are still pretty pessimistic. I do think GPT-5’s failures indicate some systemic misconceptions (see, again, items #1–7), but humans can get awfully, even terrifyingly, far operating in a purely LH mode of cognition. We can have genuinely passed the inflection point on the logistic curve, after which we will see fewer and fewer dramatic breakthroughs and more and more high-profile failures… and still have a long, long way to go.
The larger point to be made, and the place where, in my view, real unpredictability enters the picture, is that none of these developments happen in laboratory-style isolation from their reception in culture. (If there is any such thing as “laboratory-style isolation,” on which again see Personal Knowledge.) That reception, as far as I can tell, is not going well. The proliferation of “AI slop” is incredibly unpopular, even if at this point few people can reliably distinguish it from “real,” human-grade slop. I cannot imagine parents (at least in upper-middle-class contexts) long enduring the indiscriminate application of chatbot-based ed-tech products to their kids’ classrooms; Jonathan Haidt’s The Anxious Generation seems to be striking too strong a nerve for that. If (speaking extremely loosely here) the washing machine created “the problem that has no name” for mid-century housewives, thus accelerating the development of second-wave feminism, what social upheaval(s) might the washing machines of the AI field accelerate or unleash? I have no idea. And those will shape the outcomes I predict in #11 and, especially, #12. So with that (very large) asterisk placed next to them… I more or less stand by them. Rogue AI is not going to decide it needs to wipe us all out with commandeered nuclear weapons to protect its colonization of the universe (the doomsday scenario in “AI 2027”), nor will properly aligned AI usher in a drudgery-free UBI paradise (the golden-age scenario in “AI 2027”). Both of those scenarios depend on an inflationary and wrong view of AI’s capability to genuinely imitate human intelligence. Instead, the real damage will be done by those purportedly racing to prevent the first and usher in the second, who are on public record as having zero clue what their actual end goal or desired end point is. They seem to imagine it will go on forever and ever, growth without end. But Stein’s Law is undefeated, because it is in fact a restatement of the second law of thermodynamics: “What cannot go on forever must eventually stop.”