During a dinner put on by my high school’s IB program several years ago, I recall standing up to the microphone following some remarks delivered by classmates. I joked that I feared following the general pattern of every HS history seminar, “Increasingly clever ways of saying the same thing slightly differently until nothing is said at all.” Sitting down to write this year’s post, I have this remark in mind because many of the ideas and events I feel compelled to reflect on are extensions of what I jotted down previously. This is what I intended writing begets more writing. I expect the majority of this year’s essay to draw out one puny observation that I admit did not appreciate the significance of in December 2022 that I had no idea what’s in store for “what novel forms of collective mind OpenAI revealed are advancing at breakneck speed.” I recognize the danger of revisiting the same topics with slightly different weights. The 10,000 word pastiches I have planned for the remaining decade are like the mad British painter, depicting his cats with stunning clarity, until his senses gradually then suddenly withered away, and in the process anticipated a new genre of art.
Thanks for reading The Moving Train Book Club! Subscribe for free to receive new posts and support my work.
Will 12 Tweets follow this path? (Above) An obligatory Christmas themed image from the mad British painter referenced above. (Below) Louis Wain’s trajectory of paintings here demonstrates the meaning of Model Land, “through art to know what nature is not.” Beware of becoming a permanent resident because poor souls who cannot escape often grow mad.
I embrace the danger because that is in fact my goal for what I hoped to achieve in these posts. Recall the Picasso line from last year, imparting unexpected views on Model Land more familiar to economists, “Art and nature are different, for it is through nature I know what art is not.” Alternatively, the slogan that resonates in the moment is a Japanese proverb, “Mistake after mistake compounds to the sum of all human knowledge.” I can think of no better phrase to hang in OpenAI’s offices, one that captures the process as well as significance of the workable collective mind that can be summoned on the slightest whim. It’s a vindication of the Rosa framing of modernity as accessible, attainable, and available. Yet it should be emphasized when staring at the GPT-4 masterpiece, Shannon’simpossible dream, that we know through art what nature is not. The linked historical review of LLM’s asks the obvious question: what is the point of constructing a generative masterpiece if the aesthetic judgements necessary to appreciate it atrophy?
Though I am usually content leaving my personal reflections separate from my social commentary, I believe that this post will test the limits of that inclination. Readers of this blog know that I am frequently drawn to cases of successful failure, chiefly American power, hyperglobalization, and modernity.1 I’d argue that successful failures are rooted in a common myth, that of the Gyges ring— if presented with a device which grants the wearer with Rosas’ reach, do you have the ability to refuse?
Tragically, I wrote last year’s installment of this post, I have never been able to exercise the mental discipline to consistently answer no. And that makes me like the waters I swim in, a successful failure. Entering my teens on the heels of the Great Financial Crisis, if you asked me about my professional and educational goals, I would have said to study economics at LSE or Wharton with interspersed attempts to make sense of the behemoth that was reshaping the world economy. The world spirit, I surmised, is read in Chinese. I did both of those things precisely as intended, but the effects did not accord with what I envisioned.
That should be interpreted less as an admission of the limitations on my self-making power than as a point of emphasis. The polycrisis is inherently perspectival. I will never hesitate to share my views on the perplexing American economy but rarely will I transgress the successful failure framing. To paraphrase a superb Menand review on Gibbon, I recognize that this is my “left testicle”, the personal insecurities which influence the stories I choose to tell. Stories are precisely that—choice compilations of narratives to make sense of reality as we experience it. But what happens to a story after it is hard-pressed, another terrific New Yorker review pressed? Wittgenstein’s tormented soul is the appropriate messenger for this insight as the sense of direction and composure loses grip. Tragically for me and comically for thinkers like Wittgenstein, When We Cease to Understand the World, the quantum world where we are continually surprised by our discoveries and defined in opposition to the hydraulic inventions governed by classical dynamics, remains as relevant in 2023 as 2022.
I do not yet see the higher plane of comedy. The world is simply changing too fast. But the AI world as viewed from the ground floor—the cigar shop—is fed to me with a dose of tragedy. The comedians are the masters of the universe, who due to their position at the pinnacle escape salvage which Anna Tsing believes defines capitalism, interpret LLM variants of AI as a dud. While it is nice to have a shiny tool that is willing to tackle the legacy code even the pointiest members of the software priesthood were afraid to touch, the products currently offered simply do not have capabilities which exceed the institutional know-how available to them. In fact, they go further than this. LLM’s are extremely poor imitations of the polyphonic discursive rhythms which govern how institutions think.2 Unlike these institutions, they have no world-making model. Call it what you like: stochastic parroting, spicy autocorrect, but it’s not thinking.
The comedian shouts his well-rehearsed routine to a tech audience which preaches to the less tempered corners of the animal spirits beholden to the 10x bible. As Sam Altman tells it, AI will not simply deliver marginally better enterprise software or rejigger the Perrowian interactive complexity at the heart of financial machinery but be a boon to all of humanity. Namely, it will deliver productivity improvements which have been the dog that didn’t bark for the last 50 years in the Western world and make us “immeasurably richer”.
I have seen enough from Altman’s product to entertain the thought that he might be right, though like all tech evangelists he refuses to steelman the logistic case of resource and energy bottlenecks. That willingness is not to discount the comedic view. No version of ChatGPT short of a Kuhnian world-making paradigm shift will deliver 10x gains to $400,000/per capita leading edge firms in software and finance. But depending on the social context in which the technology is received, it could deliver something in the ballpark for the $40,000/per capita world of long tail service firms. Rarely stated this explicitly,3 the Altman solution to the productivity puzzle is to take the trend depicted in this chart as a given and up the dosage of the medicine. In the future, the top programming language will be English and contingent on how systems improve on key hellaswag metric, it may not be particularly important to write at high level of precision to attain desired output on puny signal.
In a manner that was simply cruel for out-of-touch Clintonites to retort to industrial working classes who bore the brunt of austerity in the 90’s, we really can all become coders now. The status distinctions which once separated those who read binary from those who don’t have all but evaporated, prompting The New Yorker’s eulogy to coding. The Tsing/Rostow foil/fusion point from last year (tweet #9) resonates here. In the ruins of progress and on the cusp of maturity lie the promising avenue of new opportunities yet realized. Software may have eaten the world, but deployed in the hands of folks who possess all their senses, we just might build the tools we need.
Mutatis mutandis the power embedded in software tools. Normal people when suddenly granted the gift of sight can possibly translate the polyphonies they hear that the software people never attempted to build because they were too busy playing.
Other evidence points in the direction that is often suspected. Just because the spell books have been released does not mean the AI divide has been breached. All else being equal, early adopters are younger, highly educated, and employed at larger firms. The ordinal society, Fourcade’s warning, is the familiar hourglass quandary under a new guise.
The hope is as suggested by the Tsing/Rostow message. On the one hand, policymakers face a series of ruins to navigate. On the other hand, even absent action from policymakers, I suspect qualified people are embracing possibilities in all kinds of places that would have not attracted in prior era under old career ladder(s). The canal-rhizome has replaced the tree, and consequently perfect ordinalization is an interrupted script. The impoverished sophisticate exist in the spaces in between.
That is the positive spin to the gothic visions of Turchin’s elite overproduction thesis that are more commonly imagined. Rather than succumb to siren song of rump intellectuals, fallen professionals may answer the intellectual challenge to make the economy more productive precisely in the places that have long perplexed distant elites. The difficulty is the attraction, a companion to the frustrations often faced in the developing world. Development is a journey for all.4 They will succeed, infusing the economy with new dynamism that is already observed in latest prints for business establishments. Notably, in the U.S., some of this long tail is contained in regions which have been historically excluded, such as the southeast.
We should celebrate the success aspect of successful failure. The alternative is eminently clear in cases like Italy, Greece, and even the sluggish recovery of the 2010’s when Big Fiscal was kept unnecessarily restricted. Unnecessary restrictions, self-incurred immaturity, is another theme that unites. I am now painfully aware that the mind-numbing office computer tasks I spent hours on in the mid-2010’s could be executed in a few seconds via script by a friend at Google. The essence of what Roberto Unger intended by inclusive vanguardism is that those tools are no longer the exclusive right of those who have access to the deep reserves of institutional knowledge present within the vanguard.
Of course, Unger’s principle is a moving target. OpenAI has ushered what remains the early days of the AI movement, and whether the new vanguard will apply inclusivity reflexively to itself remains an open question. I am hopeful, even if the corporate-minded shift of OpenAI’s board, symbolized by Summers, following the Altman debacle indicates a trajectory of maturity technologies towards insularity.
Failure mode should also be guarded against. I see one potential tragedy as this. Rather than take productivity gains to partially substitute the technical for the creative, the frenetic pace of advancement overwhelms all else leading us to be less patient with creativity. I started this post by stating tread carefully when unleashing full range of imaginative states unhinged from nature. Everything, Everywhere, All at Oncepresents a Jobu Tupaki-sized problem for people like Joy, representative all of us, who are finding their way. The process cosmology of becoming, Yaqing Qin’s enlightening jargon toward a unified theory of the present, from the transformation of finance, covered in Wansleblen’s incorporated comparisons, to the search of solids from within are crucially mediated in the film through Chinese-American voices.
To be sure, the reverse could just as easily be true. It is frenetic rationality which must be disciplined if we are to pull off the high wire balancing act on the edge of chaos. How is imagination deployed: in service of the model5 or an intuitive judgment of when to periodically hatch an escape? Subconsciously organizing my mental states on latter basis, I side with GK Chesterton in an perennially relevant critique of nearly all social science6, The Medical Mistake. Chesterton’s description resonated— if limited to ordering bits of information into sequentially precise locations without poetic release, I’d grow mad. Anecdotally but I think demonstrates the phenomenon well, I can’t tell you the number of Trumpers I’ve encountered in the past year who are professionally governed by iron cages of rationality but because they don’t know how to be primitive, they embrace barbarism. The more primitive you are, the less primitive you become.
The post-GFC decline of the humanities suggests the U.S. has not heeded Gombrich’s lesson. People respond to elite-structured incentives extraordinarily well, and there is nothing more boring. Tied to the mast, they miss out on the unstructured play happening on the island below. The New Yorker report featuring Larry Summers prominently on the tectonic shift from humanities to the technical professions which pay paints this tragic image. Within the image, successful failure reveals a generational hue.
To the extent we, children of the GFC, are simultaneously entering the computer’s world of scale, an alchemy Weil interpreted transforms that which is human to that which is inhuman, we may be rejoicing in instruments of our oppression, which if constituted differently could just as easily liberate. “To serve man” has two meanings, an episode of The Twilight Zone once observed. Thus, we have reached the Heideggerian indeterminacy of the qualia7 which shape technology’s course. I’d be more confident in that course if men like Summers8 did not mimic the 9-foot alien’s intent when they utter “to serve man”.
Larry Summers Mark Zuckerberg Elon Musk Marc Andreeson The towering aliens from the Twilight Zone reading in a telepathic language (he doesn’t require human communication) all the ways he will “serve man” [in a soup].
I should note that the overwhelming focus on this essay’s preface is on possibilities and prospects of AI’s development. I have little doubt that I will look back on this decision as a failure to look beyond my crumbling fortress. Wittgenstein believed that looking depends on noticing “the hurly burly”, the breadth of the background. If the 2021 commentary was weighted overwhelmingly to climate change, 2022 as the rise of the rest and resulting shifts in geopolitics which remain mostly undigested, and 2023 as the social context of creativity, only 2021 is appropriately weighted. We begin an investigation into these social forces in a direction that has long inspired reflection, China.
The First Movement: A Question of Sight
On the first post of Christmas:
I think it’s appropriate to begin on this topic because as Tooze emphasized elsewhere, we cannot reflect on the significance of 2023 without noting that in the first 4 months of the year the top brass of American foreign policy and defense circles were seriously preparing for war with China. My question from last year, it turns out, did not stay confined to my niche corner of the cozyweb. Absent the trust that underpins open access orders, many of us are currently wondering will it sag like a heavy load or will it explode? Fortunately, we sided on the former with Yellen visiting China and Xi visiting SF during the latter half of the year.
Tooze’s historical instincts are correct if they are summoned to imply that this moment will have Cuban Missile Crisis levels of weighting for future historians relieved from the daily humdrum. Attention is a strange thing, I emphasized last year. What prevents us from seeing the 800 pound gorilla when we are distracted directing our gaze at juggling balls? It’s possible that a few of us read the Ed Luce column, but it wasn’t quite the DEFCON3 alarm it should have registered as. The nature of of attention in the Global Risk Society is no single factor will ever be weighted appropriately and sense of imbalance leaves the observer constantly spinning between overload, vertigo, dizzyiness, and trance. Some of us, including I suspect many in the White House, are considerably more fixated on the drift of the GOP to the point they were banging the drum to invade Mexico. The tragedy for structuring global risk societies is though useful knowledge agglomerates everyday by Hayekian magic from ignorance, wisdom never results as the output from divided attentional states. The peace interest, Frederick the Great surmised, requires some entity which “thinks with one brain”.
I am troubled by the long game of telephone that is necessary to transmit policy knowledge into actual policy because it seems to me that the people who devise the instruments are not the people who should be entrusted with any matters bordering on sub specie aeternitatis. It’s not that those who live in policy worlds should be trusted with nothing at all. In sub specie communitatis, groupings of people which fall short of all of humanity and are constrained to the present, policy bullshitters often do quite well because, for whatever reason, the game of telephone is significantly shortened. PolicyTensor recounted to me that he had solid reason to believe that his Biden Boom memo directly influenced the White House.
The refrain articulated earlier should be repeated. The success aspect of failure should be celebrated for it offers the possibility of more general policy learning. We have escaped the exasperated stance of Keynes who watched in horror as European elites doubled the human catastrophe of war with fiscal fecklessness. Small consolation, I know, but at the very least, U.S. elites did not double the mental and spiritual crisis of pandemic with austerity. Consequently, those of us who feel so inclined, like Robert Hockett, can celebrate 6 percent quarterly growth prints and the return of frying pan charts.
Neither Hockett nor PolicyTensor are mentioned at random. Hockett, in particular, is a direct advisor to the members of Congress who have the patience between fundraising calls, notably Ro Khanna. The policymaker, who paraphrasing Orwell does not possess malign intent but limited attention and is therefore unteachable, takes the shortened formulation of the advice. Manufacturing is good, because it will incubate communities of engineering practice which will power the green transition. Operating under the narrative policy framework, they also hear voices in the air which imbue the content with meaning.
And thus, the impossible dream of technocratic liberalism of a perfect communication machine stripped of language remains elusive. Instead, what my LSE professors called the garbage can model predominates, those narrative elements which are in the background such as China is the villain which decimated US industry and caused deaths of despair, are elevated to the foreground by bullshitters who have not trained their attentions to discern fact from fiction. Their detours prove that entropy stands in the way of the Niebuhrian hope that liberalism may one day be an efficient force of history.
Yet in the ruins of impossible dreams clarify the actual task at hand. If any of us find ourselves in a position to give policy advice, we ought to first be under no illusions that policy can be transmitted without narrative and to construct the narrative background carefully to ensure our words cannot be twisted into creations that horrify.
Nor is it the case that liberalism per se stands in the way of efficiency, but rather the fundamental obstacle of the perfect communication machine. Kaiser Kuo’s observation in the aftermath of the balloon incident that he can assemble a convention-sized hall of reputable academics, think tank specialists, and gifted media talents who side with Jessica Chen-Weiss, articulated powerfully on the Ezra Klein Show, over the Hill/DoD view could in fact be assembled with astonishing symmetry with their Chinese counterparts. They despair over their wolf warrior diplomats just as much as our GOP congressman, and the Big Men leaders who rally them.
Their malaise is in fact greater because the political economy collision which will govern the simultaneous provincial rebalancing from investment-directed to consumer guided economy is still to be determined. The Chinese cannot yet claim the same level of policy learning vis-a-vis 80’s/90’s Japan as the U.S. can claim over the arc from the GFC to COVID-19, from Crashed to Shutdown.
On both sides, the sense is that relationship would be on firmer footing if we skipped the uninvited members altogether. The fundamental division in world politics is not between world powers but the cleavage between the Ionians, those motivated in the Spinozian tradition to understand, and the Athenians, those granted with power but philosophical zombies devoid of this desire. Boy wonder, Jake Sullivan, fits the bill with astonishing words in The New Yorker.
If we felt so inclined to brainstorm, we would conceivably come up with two possible channels for autotomy, leaving the Athenian half for devouring by world and with the second make good our escape. The first is the Platonic ideal. Simply elevate expertise to the point where efficiency of the machine is an irrelevant concern. This solution is not as prescriptive as it suggests for the question is how to elevate. Tradeoffs of attention are fundamental since spending large sums of time with scholarly works and behind computer screens interrogating reality crowd out cultivating elite networks that shape it. For this reason, the Plantonic ideal is often only realized posthumously, as is the case for Marx. And from the grave, the Athenians find ways to subvert it. The New Yorker’s cartoon portrayal will be perennially true.
The second solution is operationally simpler, to deprive the Athenians of the powers of rhetoric they are ill-equipped to wield. In that scenario, the void would be filled with the populist ideal often associated with the American presidency. That to slew the ship of the state, leave the elites comfortably along the margins. If they are kept inside, they may mimic the aliens discussed in the introduction, steering more effectively initially but purely serving their kind in later stages of the game. What is needed instead is a president who shares the DNA of the people, possessing their moral judgments to govern uncertainties which cannot be calculated and are entirely novel.
I do not believe the populist temptation to be fruitful either. You only need to know that it was Spock who steered the ship on the brink of madness after a COVID-like episode in Is there No Truth in Beauty?9In our world, you only need to refer as far as opinion polls of the percentage of the American public which favored deploying nuclear weapons against the barbaric Russians to expose the populist strain as myth.
Though, there is a productive logic to my imagined exercise with Chinese colleagues because it demonstrates for good that there is no such thing as exit. Only voice, better communication with Athenian elites and the public, offers the path to restoring stability in the relationship. The progressive logic extends to the chips escalation itself. To be sure, not progressive for us10, but for them in that it will engender creativity in a manner that is familiar to many of the gargantuan chemical, engineering, and scientific efforts of prior world schisms. Sub specie communitatis. To the extent they hold up half the sky in the green transition, with miraculous growth in the EV auto sector noted this year in the FT, their contributions may reveal a shade of sub specie humanitatis, which can then ascend to sub specie aeternitatis under the right conditions.
On the second post of Christmas:
I’m developing a rhythm in these posts with the first tweet offering some reflection on the philosophy of history inspired from first part of the year, trending toward terminus (“explode”) or muddling (“heavy load”) followed by a confession. A word of caution is in order to this confession prompted by the image of Hanania flipping burgers. The extended metaphors of eating bitterness, sent-down youth, and myopic intellectuals are not to be interpreted literally. The relation between the global tumult of the 60’s, literally the world turned upside down, and today is about as strong as total war metaphors for the IRA response when all that is on offer is a versatile but wholly unfit for purpose Swiss army knife.
Last year, I admitted that I was at times a tight money liberal; now I will write that I am occasionally drawn to Maoist instincts. And for the same reason— from the perspective a distant observer, a progressive logic to reaction comes into view. Of course, neither thought must be entertained for more than a few seconds if you are to take seriously first do no harm. The corollary is to hate precisely, to reject “no Mao, no modern China” bromides I frequently heard from college-educated young people I sparked conversations with on travels around China.
The jarring reality I am coming to terms with is that we might be settling on Maoist conditions, finding wannabe intellectuals to send down into the world, in the absence of a charismatic leader which revives the mass line and completely unhinges the social order. Which is to say that the crisis of the university as something more than human capital development facilities is inviting me into an exercise of what exactly I hate about Maoism. The strong position is that sent-down youth, abstracted as superbly talented individuals or even capable ones cleaning latrines and other tasks that do not accord with those talents, is an affront to the social order. I propose a weak position inspired not by conservative beliefs but mentality of desiring that which most instinctively keep their distance from. In essence, the young Edmond Burke recalling the exhilaration as the waters engulfed his boyhood home. I find something sublime about having some extended period of time where a group of people are forced into this position because the funding and positions which existed in a prior era have disappeared.
The old social order, according to Huxley, is inherited from the Middle Ages. Naturally, people fall along a spectrum for aggressive and contemplative tendencies. Those that fall along the middle of the spectrum pose no threat to the social order and could be employed productively as serfs. But the other groups require institutions to deploy their energies productively, monastery-universities in the case of the contemplatives and knightly orders in the case of the aggressives.
For a society whose modus operandi was homeostasis, this exit solution provided precisely the release that was needed. But to someone like Mao who was seeking to exit stasis, the social order was precisely the problem. Moreover, the proportions of contemplatives to aggressives is never constant. As societies decay, the balance is upset in either direction. This was likely the context of Mao’s remark during the Yan’an period of the stinking ninth that “intellectuals are more useless than shit”. He of course did not mean that intellectual inquiry is more useless than shit. He was mimicking financial professionals like Russell Napier or Gary Stevenson today when they decry the uselessness of the academic monastery’s models, which have developed in a zombie world independent of reality and incentives.
If the point is to know, the contemplatives must not merely imagine. The process is backwards. They must first see, then they can do, and only then can they imagine. With pre-revolutionary China facing grinding poverty, the first loop of learning is the most essential, “The rational is reliable because it has its source in sense perceptions, otherwise it would be like a tree without roots, subjective, self-engendered. Sequence in cognition, perceptual experience comes first.”
In advanced stages of decay beset by Buddenbrook dynamics, the sequencing is all off, and man is mismeasured with imaginative capacities misaligned with actual senses. That impoverished sophisticate taken from the monasteries, making the trek to Yan’an, cannot drive development unless they are made uncomfortable, to drive them to step 1 of Buddenbrooks dynamics obsessed with basic security. Survival trains them in the arts of noticing. Seeing in the service of security, systemic vulnerability, makes it so that they possess not only the mental models of the world— in theory possible under step 0 pure imagination unhinged from sight. But also the second component of intelligence, the motivation to realize the mental models into action.
The highest stage in the third loop of learning, counterfactual imagining, is elusive but it is not a stage which is separate from action but immanent to it. Mao may have sought out to smash the four olds, but not without first imbibing their lessons, “Learning without praxis is a waste, praxis without learning is perilous. [Alternate translation] cognitive discretion without executive is impotent, executive discretion without cognitive is arbitrary.” There is a deep point suggested by Huxley critiquing the two cultures discourse that was all the rage in Britain, another society that was obsessed by the ideology of declinism. The discourse was missing the forest in the trees, the myopic prophet interpreted, because neither left nor right-brained intellectual activity will reverse the course if those activities are not coupled with activity in the world, to follow Spinoza and unify the mind-body to get closer to God. Exploration and exploitation is not a tradeoff. Only in exploitation can you realize the fruits of exploration. Invent and wander.
By coincidence, Huxley intuited the cutting-edge cognitive science which is currently engaging the AI researchers at U.K.-based DeepMind. Fixated on fusing the two cultures, techne and episteme, as that which frustrates, technical and social science expertise develop increasingly clever means to unimproved ends. Intricate, multi-modal models often appeal but they further aggravate Chesterton’s medical mistake without the missing element. The second component of what makes a model intelligent, motivation, is much more necessary as the substrate of history.
Presumably for corporate clients seduced by the prospect, the holy grail of DeepMind’s research is artificial capable intelligence (ACI). That is not to simply devise series of rules as defined within the model, but escaping the model to define the series of tasks armies of business associates working on all facets of product development organize to achieve. Rule-based governance occurs first in the Mencius phase of development. Open-ended task generation is what it takes to incubate a community of engineering practice on Tsai & Chang’s terms. Even accounting for his boastful sloganeering, Mao exceeded those objectives beyond his wildest expectations. He developed the Musk strategy of overpromise before you can deliver, and like Musk, his company, China Inc., eventually lived up to the hype. A certain amount of madness is required to invent the future.
With enough intellectuals driven into the world, they might contribute to closing the most important chart in the world today (see below). Proceeding through all three loops of learning where Huxley was confined to imagination, they echo his message—what miseries of (post) Fordist man be avoided while still detected at this early stage. Men of Huxley’s 0.5% which inspired the alpha double plus stratum in Brave New World like Larry Summers were so far above, they could not imagine those miseries and they deindustrialized. Coupled with the deemphasis of the humanities, tight money liberalism is now recognized as an epic mistake which perversely accelerated the decline it sought to reverse.
Both policy courses have been so thoroughly rejected that the alpha double pluses risk irrelevance. The motivation humanities can engender are better wielded in our hands. Given that our task is different from theirs, we might not need the same number of humanities professors which meet the objective of elite reproduction and social stasis that we needed in 1980. The world of 2023 is simply too interesting to teach students in the artificial environment of the university-monastery built for statis when the upheaval of the moment demands that we find our own students on the many planes which offer moral equivalents of Yan’an, answering the guerilla spirit which developed both countries. Just as I think my country has reached the end of the road and feel the urge to retreat into art, I realize we have not yet started. I am pushed into starting afresh, discovering that the scaffolding of personal as well as national development is never finished.
On the third post of Christmas:
Whether I have Mao or Andreesen in mind—and indeed others who will be mentioned soon from Bezos to Musk—to hate precisely is the question that drives my thinking. I have already noted that there is much I find useful about their views. Andreesen’s descriptions of getting software/hardware bottlenecks off the ground in the 90’s are enlightening. The aspect which I hate about them is simply the genius problem.
The Genius Problem is the oldest myth that there are a class of men who live up in the clouds. Any attempt to regulate them from the ground will be ineffectual, so they must be free to drop their innovations from on high even if/especially (depending on left/right position) their wrath engenders divine fury mortals are ill-equipped to handle. They are the exponential men, living in exponential time, and their world(s) collide irreparably with ordinary linear time. The simple formulation “without Mao, no modern China,” can be adapted to any man who ushers the second kind of time. Without Andreeson (and his Netscape browser), no modernity.
The connection to time is fundamental as it seeks to replace the one God clock-centered universe with an aristocracy of talent. The claim to legitimacy of the creative class is to point to inventions as proof of their awesome divinity. The Newtonian world un univers horloge, un dieu horloger one clock universe, one god had been smashed as a result of their machinations.
The Newtonian world was much too ordered, and in any case they have tactically released a process, dubbed effective acceleration, which cannot simply be undone. The proposal for those in the vast majority who live in linear time is to state clearly that the cloud-length relationship is over. As the late Bruno Latour might have said, they must be brought down to Earth.
The hope is when they are brought down to Earth they will discover the insights which their distance allowed them to long neglect. The most fundamental decision, Jonathan Taplin wrote on the history and future of oligarchic technocracy, is those who choose to live life as humans and those who choose to live life as machines. As another person who is capable of insight but is rarely mistaken for genius, Michael Lewis, observing SBF , a new type of human has been born. Somebody who is so hooked on the dopamine drip of information— regardless of the beauty and meaning of the content, it could be as meaningless as a shitcoin or first-person shooter game—that he’d prefer internet access under real deprivation over awesome real abundance but without information.
In a sense, Andreeson has already spent enough time on Earth to have a peek into the future of what he presents as his solution to “reality deprivation”. He says in our cloud worlds: we have extraordinary access through Rosa’s reach, the monstrous sovereignty platform capitalism provides, human felicity attained. Others do not have the luxury of accessing real versions of human felicity attained. But the dream I envisioned in 90’s, that a person far outside from academia or policy who happened to have a neural wiring which gives him a massive dopamine surge when listening to a Tooze lecture could do so on demand. Most people do not have this strange neural wiring, but however their synapses are arranged we will have algorithms to identify and provide content with frictionless regularity.
I am left wondering how such an insightful figure, deprived of the feeling sense, can see this as desirable, and not a realization of Huxley’s ominous warning. That is the second revision to Brave New World, prompted by his move to LA and first-hand encounters with post-war prosperity, The Grand Hotel Abyss. The Ultimate Revolution is downstream of Hume, that the ruled will effortlessly resign their passions to the technocratic oligarchs, “that man can be made to enjoy conditions, that, by any reasonable standard, he ought not to enjoy.”
The course of technology as indeterminant returns as Huxley exits stage left. The tools the oligarchs have provided, in the process shattering the one-clock universe, could liberate if the myth that select few have divine power for creativity others lack is upended. Virtuosity, tail-end talent, is clearly observed in the wild. Yet, to paraphrase another modern poet who relied on his feeling sense, David Bowie, virtuosity is uninteresting.11 The everyday use of the language gene is the essence of what makes creativity innate, and in my view more interesting than even Mozart.
While the right insists on virtuosity, that creativity will always be confined, the left suggests language points to the place where humanity ought to live. Whether we can upgrade that innate sense should be the objective every person has on this Earth to better acclimate their thinking-feeling sense, and acting to align that sense with the world. The awesome creative power embedded within language, Spock intuited as he mind-melded with the Medusa, is that which will allow humanity to “return to the small things we are masters of”. Until the message from the concluding line of that poem is understood by the oligarchs, the vastest of things we may not learn guided by love, they will find themselves in position of having their inventions jeopardized. “How small man is/ yet in his mind how great/ believing he is master of all things/but scarcely can command anything,” Edmund Burke.
On the fourth post of Christmas:
The content in the above excerpt, Jenkin’s interpretation to the Poussin painting of the same name, is chiefly philosophical. The poetic of ruins, according to the assessment of an author whose book project, a philosophy of computing from Leibniz I will be monitoring, was a motif for thinking about history that gave birth to modern discourse during post-revolutionary France. In manner that is simply not true for any period before, their issues are immediately recognizable to any educated person today without special historical training. The rhyming is also not mathematical in the vein of chaos theory i.e 50,000 years of chaos but direct. Their history, particularly the coincident events of the end of the Napoleonic Wars with the climatic shock of 1815 is our history, the setting of the Shelleys’ Frankenstein and Prometheus Unbound.12
Trained in a new discipline growing out of AI ethics and effective altruism at Cambridge, mathematical philosophy, Burkowski writes of Mercier’s Paris on terms diametrically composed theorists from their oases at UC Santa Cruz and elsewhere would appreciate:
Cut off from its past and with no discernible future, post-revolutionary Paris was drifting aimlessly in a state of generalized confusion caught between fantasies of regeneration and prospects of looming destruction. Construction inevitably seemed to spell future destruction, linking ruins, past and future in a strange and seemingly inexorable “poetic of ruins” before which all civilizations were called to disappear. Pure historicity seemed to have reached its logical conclusion.
But it’s the personal resonance that ultimately shines through, the intimacy Powell’s portraits intended to summon. I often feel this way cut from the past and as a result drifting aimlessly. To be sure, with fantasies of revival conjured up alongside horrors of general decay. With neither the fantasies nor the horrors tethered to solids. Regarding the split, I consider the outward tragedies I observe. My grandfather used to recite a line relevant to the abject denial of U.S. elites puzzled by those who fail to keep up with the pace of the dance. Burn-Murdoch’s trademarked data journalism for the FT documents just how unusual the U.S. is among the developed world for under-40 mortality. Presumably drawing from Rostow, he lectured me that growing up is a race between maturity and disaster. The U.S. makes a policy choice when they construe disaster to mean death, prison, or being thrown out on the streets.
Dovetailing with the mission of Antony’s channel, I am open to defining what those solids might be, but for the moment I admit that I am simply puzzled. I am left only with the smoking gun of older generation’s descriptions of daily life. Prior to Bowling Alone, the solids were constantly available, so available that nobody even took care to study them, part of the hurly burly. Under liquid modernity, there can be little doubt they went dormant.
How pitiful is our enforced return to those small things we are masters of is precisely the romantic sprit which drives me with mixed results. Unless you are prepared to live with pitiful results, I am not sure that I would recommend becoming too enamored by this idea. We will see shortly that the path of artist nations and people is toward ruin. Ruin that is often heroic and noble from the the eternal, but quixotic from the present.
Yet the belief is so foundational I see no other way to proceed. The influence originates from Bertrand Russell in a short dated book, titled The Problem of China, arguing with Orientalist gaze that restless Westerners always seeking domination must reinvigorate their spirit with the way of the Dao. And the artist Chinese, if they hope to survive in this cruel world, and not have wickedly unwise Japanese and Europeans trample on their civilizational inheritance must adopt the scientific method to unleash general improvement of technology and social organization. The modernized version of the same insight hails from digital media theorist Douglas Rushkoff’s reliance on the concept of wetiko that we restless Westerners cannibalize the life force and bizarrely confuse it as morally upright conduct. Cory Doctorow, whose book I added to the list for next year, has an intensely appealing theory of internet decay, a novel accelerated variant of more familiar political form but interacts with it to generate the revolt of the public.
The indigenous perspectives these authors summon may seem far removed from the advanced “modern” civilization most of us enjoy every day. Though, the reason for inviting them is to remind that we have never been modern. On the grand tour of The Material World, along the same path followed by Latour reading newspaper headlines, Conway searches out the solid stories he only saw in liquid form on spreadsheets as a materials analyst. The binary he found regarding the electrostates, the only places in the world which can provide the crucial minerals needed on the speed and scale to fuel the green transition, is classic Hayekian genius/ Hayekian idiot. The vastest things we may not learn, the gargantuan projects have been so alarmingly productive that they are nearly self-perpetuating. Thus, engendering the illusion that neither human knowledge nor skill is required from the vantage point of the vast majority who will never observe is a marvelous testament to homo faber.
The Hayekian idiot is the neglect of Polanyi. The extraction process delivers the moderns with benefits which will be inserted into their EV's and encase their software but they will leave us with open veins. The moderns and their conspirators, most recently Javier Milei but travelling along a long line from Bolsonaro to Potosi, blame the primitives for thwarting progress, and refusing to engage in those processes, Russell’s scientific method, which perfect the chaos otherwise found in nature.
The minority enlightened view amongst the moderns is that it is absurd to blame the victims for their misery. They are giving us vital information that, in all our Maoist enthusiasm to move mountains, we may have aggravated rather than solved the 5000 year long duree of chaos. The modern proper tell the primitives that unfortunately we have no choice. If the music is to continue, to say nothing of the dance, we must move one last mountain under the hope that no more mountains will moved be and the impossible dream of a cybernetic energy system, net zero, will be met. The primitives’ skepticism that the psychological conditions, the conquest of happiness, do not exist which moderate the escalation of the mad capitalist dance is warranted. Jeremy Grantham, another British elite following in Russell’s footsteps, said it best, no MBA would ever be minted if they considered the planetary constraint and refused to solve this bargain:
The hourglass on Conway’s cover may run out. Is Russell’s worry, anticipating the collapse of the climate system indirectly via thermodynamics, of riotous living followed by destitution in New Hopes for a Changing World prophecy? I confess that this thought is not too far in the background every time I pick up new tech gadgets for the Christmas season.
To relieve that melancholy thought, let’s return to the second paragraph of the initial tweet with a picture of Huxley with his cat:
The Second Movement: A Question of Action
On the fifth post of Christmas:
Marcel Proust once said that nine out of ten problems that visit intelligent men stem from their very intelligence. The same holds for intelligent societies they inhabit. We have already discussed one possible explanation for the crisis of the imaginary, what we might call the constraining problem. In the vein of the medical mistake and the problem of motivation DeepMind is currently investigating, increasingly intricate multimodal modals where colossal energies must be deployed simply to determine the relation between elements drains the will to act. In the language of cybernetics, we are governed by an empire of normality such that the system is designed with variety attenuators. Too much variety is ungovernable.
I have intentionally constructed these posts as invitations to playfully explore, not to instruct. We therefore might also consider the alternate problem of empowerment. Under this view, the crisis of governability is precisely that the variety attenuators no longer function as they once did.
Adam Curtis sets the stage in his opening to Hypernormalization, noting that certain forces of individualism, unleashed in the 60’s initially on the margins in Greenwich village, would foreshadow our reality in the age of the smartphone. Attempting a unifying nonsense narrative by weaving in Jiang Qing, Curtis is a multi-dimensional neocon/neo-Maoist. He naturally views the artistic retreat from the real to the boundless world of the self as the problem.13 Interestingly, as I am sure he is aware, he is an active participant in the expressive refinement over the political void which characterizes hyperpolitics he diagnoses. I am too.
For people with these artistic sensibilities, the inclination is to seek out the sublime. The most sadistic, but also purest, expression is in the terrifying psychology of Ernst Junger, in a reactionary twist to Spinoza’s thought to fuse the mind-body with the war machine. It is medicine that is worse than the disease.
Indeed, I would submit that the full range of theorists from Mulgan, to neocons and even neo-Maoists are falling into the trap set by the medical mistake framing. The neocon, whose modern guise is neither Junger nor futurist poetry but Ross Douhat and Marc Andreeson, amply documents the elements which paint a picture in illustrious detail of the decadent society. Douhat then proposes that whether we like it or not the solution people settle on is to seek the sublime in religious demonstrations. Other theorists with opposed views, notably Tara Burton, note religiosity is so deep that it is present even most secular segments of public life. Hence, the Great Awokening die-in protests resemble the 19th century gatherings.
An alternative to the world Douhat paints is available, but it is philosophically terrifying because it entails a partial departure from rationality of the kind modelled on the medical profession. Navigators of polycrisis must know the medicine before they have fully observed the disease. Actually practicing doctors far removed from med school know this to be true, emphasized in Gladwell’s Blink. Governance requires kicking away the ladder of rules which structure decisions for those who are not yet sufficiently trained to make sense of what they see. Ontologically, this means to cross the historic 0.1 second barrier to that which can be observed and fully explained rationally, and intuitive judgments humans must accept if they are to make themselves useful in complex environments.
Exercising that wherewithal, the descent into decadence is easily reversed. If we look on something with the appropriate perspective, an entirely ordinary thing, the cigar shop, will inspire awe in a manner that relaxes decadence. The puzzle is that managing only our cigar shop without considering other cigar shops deadens our senses. Inspired from Mary Douglas’ culture theory, one of my most mediocre models of the year was to imagine a 2x2 grids of responses to capitalist modernity. Under the decadent society, too many of us are managers, those who accept capitalist modernity but do not appreciate it, and too few are philosophers, those who appreciate it but do not accept it. The onus is to revive the ground treaded by our 19th century counterparts, as Marshall Berman writes:
If we listen closely to 20th-century 21st century writers and thinkers..to a century ago... [even more for us than for Berman writing in 1982] radical flattening of perspective and shrinkage of imaginative range. Our 19th century thinkers were simultaneously enthusiasts and enemies of modern life, wrestling inexhaustibly with its ambiguities.
In one of Curtis’ earlier films, he traces the origins of technocracy to the revolutionary spirit—leaps!— of the engineer’s plot in the early Soviet Union, recovering an interview from one of the surviving descendants. Extraordinarily, they anticipated the potentially transformative impact of systems that would become computers. Equipped with centaur ability, Trotsky surmised, any ordinary person could become an Aristotle, Goethe, or Marx—and beyond this new peaks will rise.
As Venkatesh Rao reflected in one of his essential posts, simply having access to Google search for 20 years has granted abilities which many interpret as superintelligence when in reality my raw cognitive ability does not exceed that of my grandfather. “Men resemble their times more than they do their fathers,” an Arab proverb that has never been more true than this moment. A whole new category of chess prodigy is being trained on observing AI’s which exceed human skill who have the mental intuitions of hundreds of years of purely human acquired skill. Artificial intelligence is simply time compressed into bits, Rao concludes.
That compression will enable all of us to live more interesting lives than all others who have lived previously. In the first kind of time, they were impoverished and, in the second kind, we are potentially enriched. Potentially because you can imagine our newly acquired centaur abilities fused with the collective intelligence of all history documented on the internet going in either direction. One scenario is simply to imagine exploiting whatever position in the human organizational chain we currently occupy, exacerbating the managerial tendency. This would be the negative connation suggested by the engineer’s plot unleashed in the Soviet Union about 100 years ago.
Alternatively, another scenario I imagine is that humanity takes up the invitation to explore the complex assemblage as a result radical decline of threshold effects offered by GPT-4, unleashing Lindblom’s preceptoral vision of educated energies. These threshold effects will presumably drop even further. The revolutionary dynamic, the guardians of social science expertise know, follows these threshold effects. The only question, which necessarily differs because people in every time and place differ, is the wide range of uncertainty regarding the invisible barrier, the threshold where the revolutionary dynamic achieves its own impetus.
Harnessing only a primitive mode of vision, we are in The Quiet Before. Drawing from postcolonial independence to Arab Spring, the lesson of that book is actually summoning the extraordinary capacity of action is a different matter entirely. The cyberneticians may have had a point; variety attenuators might be necessary for modernity. We ought to proceed anyways to see if we can offer an alternative to the Carbon modernity of the division of labor, a Silicon modernity. The Hamlet principle prevails, to let doubt motivate action rather than undermine or enervate it.
On the sixth post of Christmas:
Reading Daston’s Rules in 2023 is a radicalizing experience. Though, I am open to the suggestion that I would have not been similarly radicalized if it was published in 1984 when Aristotle, the first name of the GPT-4 concept, was a figment of Jobs’ imagination. The above quote underscores that Daston’s message is to some extent a regurgitation of what has been said many times before, and it is simply the resonance in 2022-23 that is striking. Notably, the 19th century herculean achievements of calculation, delineating the social whole into one grand classification scheme that approximated computers and was the predecessor to ordinalization, discussed at length in the book is well-covered ground.
The novel element is an invitation to explore precisely what I find so revolting about the transhumanist ideas of Max More, living life as a machine. In a pattern that is familiar to the neoliberal arc from the Mont Perlin Society, the future is seen on the periphery, groups of little more than 300 people were quietly working on heterodox ideas, then conditions change for them to gain wide acceptance, but are still no less revolting.
The instinctive revulsion is orthodox because it is the natural response even from the belly of the beast neoliberal institutions. The protests of the disillusioned MIT students of the 60’s generation, archived in a Curtis-style documentary, are case and point. The modern educational system, the students correctly argued, is a vestige of industrial Britain which sought to drill the best left-brained minds with machine-like discipline, and cared little of their emotions. Their outbursts were the last, “And this is the beginning of the future— from now on all these robots feeling nothing thinking nothing, and nobody left to remind you what life exists,” Andre passionately articulated in My Dinner with Andre.
Observing the LLM revolution from the 60’s prism produces less jarring reactions than today. On the one hand, the data centers which power ChatGPT, costing roughly $700,000/day, have proven in the language of Andreeson’s manifesto and confirming Newton’s alchemic obsession, that sand can think. On the other hand, the more precise rephrasing of that boast is that sand can in engage in the kinds of stochastic parroting that humans in social groups do when they are deploying system 1 and lower levels of system 2 thinking 99 percent of the time. Daston writes that the pedagogical discipline that underpins mathematics curriculum that all learning proceeds from example to example is as true for machines as people. Human brains are more efficient, to be sure, but the lesson is heard. All you need is scale.
I recall from my primary school years when one of my teachers interviewed me about what she perceived as a mathematics gift. She told me I have not yet taught you the proper system of rules to solve these types of problems. I told her bluntly this does not make me gifted. This is an entirely ordinary creativity that most are afraid to apply over the fear of being “improper”, improvising with the systems of rules you know to explore the territory which you have not yet surveyed. Machines, I think, know how to do this in a zombified way, and it is the reason why I have never been too attracted by mathematics or computer programming, ordinary creative tools I must use to work effectively but not a source of daily inspiration.
A biology teacher later on in high school used to reiterate that there is infinite variation in life. He added not all people are similarly constituted and find divine inspiration in these disciplines. They do not simply proceed in a linear way from rules they know to rules they don’t. They possess a Picasso-like genius to generate new rules which have not yet been discovered, and I am totally blind to. Listen to either of these brilliant men and women speak of the internal qualia they experience when they open a computer terminal or struggle with a mathematical proof. It is the academic hierarchy which engenders physics envy. It’s also what makes them human.
I have only one belief that is never subject to revision. Simply because some people do not demonstrate particular talents our present society happens to value highly, nobody should have their spiritual sustenance starved, depriving them of an exploration of the talents they do possess. At the heart of much of Silicon Valley’s thinking, on display in Taplin’s narrative, is a contempt for the middle class. This amorphous mass of administrators does not do anything so valuable that it cannot be approximated by machines. So we might as well automate their jobs, replace them with six AI engineers, and place millions on the UBI dole.
Another character mentioned in the previous tweet, Ernst Junger, anticipated the psychology of indifference, via one his short stories, “Human perfection and technical perfection are incompatible. If we strive for one, we must sacrifice the other.” Note the technical and human contrast is a variant of what nature is not art is; what theory is not, theorizing is. The silent epic, Metropolis, left an impression this year with the protagonist, Freder Frederson, witnessing a double horror in the film. First, the la foule underground humans are sacrificed on the altar of Moloch, the price to be paid for technical perfection. Compounding the initial horror, the above-ground engineers who plotted against humanity are indifferent when they learn of the monster, too seduced by false Maria. Watching in 2023, the metaphor is almost too ham-handed to be true. False Maria is the prospect of artificial intelligence with none of the love for humanity the real Maria displays.
The schadenfreude presented by xrisk discourse is that machines might upend the hierarchy of the human computer legacy. That Matrix future—cutting off the Silicon Valley elite, levelling all distinctions, and simplifying all human decision sets— offers no reprieve from the horror I share with Frederson because I too would be subject to its oppression. Thus, I have stumbled onto an answer to what makes us human in an age where sand can think, again borrowed from Venkatesh Rao: “We care [as Daston would say, about the kanon of beauty], therefore we are.” I am not all certain the same can be said for our oligarchs, who imitating Junger demonstrate only disinterest.
On the seventh post of Christmas:
Given what I have interpreted as declining threshold effects enabled by LLM’s, Jordan Schneider openly wondered to Tyler Cowen this year whether we are on the cusp of the Golden Age of Genius. Though he did not explicitly take my revolutionary angle, he posed the same question as Trotsky that if the engineer’s plot succeeds in transforming humanity in precisely the way they envision,14 ordinary men will rise to level of Goethe—and for the first time in human history new heights are within reach. After all, the emerging centaur prodigies have wisdom beyond their years, and by the time they reach maturity, they will have obtained reserves of knowledge that simply would have been impossible for all their predecessors in the first kind of time. So the new heights should be expected. Or does it misunderstand the homogenizing forces of modernity which must be overcome to summon creativity?
Even with the advantages offered by augmented information environments, I can find no better answer than Goethe. His puzzle is Russell’s, how to fit the artist in the modernizer’s world. The artist is foolish in that his behaviors lead to ruin, but also holy, as Hockett likes to project from the Russian literature he enjoys via the yurodivy archetype. Across cultures, in the tradition of Erasmus’ In Praise of Folly, the religious role15 has been conventionally been filled by the only social classes which could afford to be foolish, where ruin is not equated with starvation. The elite have not filled this role at scale for several decades, and possibly two centuries. Due to mass prosperity, many ordinary folks in the rich world are afforded with pursuing the path of ruin without starvation. Can they dare to be great?
Just as I did with Deleuze last year, I will acknowledge that I do not understand the literary allusions I am summoning, so I will simply quote the relevant sections via Berman’s All that is Solid at length:
The very process of development, even as it transforms a wasteland into a thriving physical and social space, recreates the wasteland inside the developer himself. This is how the tragedy of development works…Another motive for Faust’s murder…from a collective impersonal drive that seems to be endemic to modernization the drive to create a homogenous environment, a totally modernized space in which the look and feel of the old world have disappeared without a trace. To point to this modern need is only to widen the mystery. We are bound to be in sympathy with Faust’s hatred for the closed, repressive, vicious Gothic world where we began, the world that destroyed Gretchen. But…he has already dealt the Gothic world a death blow: he has opened up a vibrant and dynamic world social system, a system oriented toward free activity, high productivity, long-distance trade and cosmopolitan commerce, abundance for all; he has cultivated a class of free and enterprising workers who love their new world, who will risk their lives for it, who are willing to pit their communal strength and spirit against threat. It is clear, then, that there is no real danger of reaction. So why is Faust threatened by even the slightest traces of the old world? Goethe unravels, with extraordinary penetration, the developers deepest fears. This old couple, like Gretchen, personify all that the best that the old world has to give. They are too old, too stubborn, maybe even too stupid, to adapt and to move; but they are beautiful people…It is their beauty and nobility that make Faust so uneasy. “My realm is endless to the eye, behind my back I hear it mocked.” He comes to feel that it is terrifying to look back, to look the old world in the face.
So much to unpack there, which I have neither the expertise nor space to do well, so we will stick to only the points that I have already mentioned. The first was not stated explicitly, but is present in Burkowski’s reflections of Mercier’s postrevolutionary dreams to transform Paris into a homogenous environment “a totally modernized [physical and social] space”, also present in Daston’s Rules. As Burkowski’s current computing project underscores, those Utopian dreams have been extended to our inner lives mediated by screens, often with perverse results. The second point of relevance to my quadrant exercise is the people Faust enlists in service of the project of capitalist modernity, a monstrous but extraordinarily effective world-bending sovereignty. The normies are even more full-throated in their defense of the project than the managers who are the architects. I despair over the fact that the most popular institution in American public life, Amazon, rests on an illusion, the perceived effortless of the one-click existence. In reality, in the physics language of The Grundisse, frictionless convenience rests on the equal and opposite velocity the backs of Amazon workers are sacrificed to Moloch.
Whether the philosophers (recall “minority view amongst the moderns”) can show the worker bees of the beauty they once knew is the hanging question of the moment. But like all prior Gretchen peasant rebellions of the past century, the sack of potatoes celebrated by Hobsbawm, Berman’s conclusion rings true: “In fairness to Faust, we must recognize how deeply Gretchen wants to be doomed…she brings it on herself.” Against self-annihilation that is equally mad and quixotic, the moderns and their conspirators win out, but, in the process the Gretchens show beauty, an authentic foil to self-development offered by the moderns. Again, the more primitive you are, the less primitive you become.
The postrevolutionary world reflected contemporaneously but at more distance than Mercier by Frederich Schiller aligns with Berman’s preoccupation with authenticity. Carbon modernity, of the kind envisioned by Mercier and related grand classification schemes pioneered in the aftermath of the French Revolution, requires a division of labor and hence the fragmentation of human wholeness. The process of development, Faust’s bargain, of course requires this fragmentation. I have written that empirically at early stages of development, the only thing that is needed is scale, and thus internal conflict between technical and human perfection, techne and episteme, was productive.
But it remains to be seen that at higher levels when the scale objective has run its course whether we will have to imitate the example of the primitives who do not ascertain the difference between human and technical perfection, and the contradiction evaporates. Echoing Fredrich the Great and articulating the impetus of the Enlightenment project that enthused both, Schiller speculates, “eventually, the dispersed faculties will have to be united in a new, integrated humanity.” And it does not end there. After the fusion has been achieved, the next step from the first two metamorphoses Dreamer and Lover, as Huxley critiqued, is the Developer:
Faust’s problems dramatized larger tensions that agitated all ….modern intellectuals would find ways to break from their isolation….The powers of his mind, in turning inward, have turned against him and turned into his prison. He is straining to find a way of the abundance of his inner life to overflow, to express itself through action in the world outside…he encounters the symbol of the Earth Spirit, and all at once, ‘I look and feel my powers growing./ As if I’d drunk new wine I’m glowing./ I feel the courage to plunge into the world,/ To bear all earthly grief, all earthly joy;/ To wrestle with the storm, to grapple and clinch/To enter the jaws of the shipwreck and to never flinch.”
The consequences and repercussions of Enlightenment are still too soon to tell.
On the eighth post of Christmas:
I have stated before and will state again that this is an invitation to explore, not instruct. So I do not want to take Cowen’s nomination of Hayek as the prophet of the AI age, announced at the LSE, totally at face value. Though, for a moment we must follow the ground he travelled in the lecture. AI carries the potential to revolutionize social sciences, he argued, merging economics, sociology, and anthropology, as we deploy AI’s to analyze the tapestry of interactions at a “a small Alaskan fishing village” and apply the principles uncovered to scale to “Belgium”. The subtext of the thought and reason why Hayek’s ghost lurks is that the observations of the individual researcher are not particularly important—the AI knows. Collective advancement proceeds through an agglomeration of individual ignorance. Can advancement still proceed under Hegel’s alchemy of quantity if nobody knows, only the AI? Cowen curiously answers in the case of crypto regulation in the affirmative.
Fundamentally, this is how AI models are trained. How does the crooked timber of humanity get translated into useful knowledge? I present to you the smiley face of shoggoth:
Hayek and Shannon’s combined insights that useful knowledge can be ascertained statistically from all of documented human experience are vindicated. The LLM revolution marks the realization of Sydney Brenner’s 2006 prediction that 2020, coinciding precisely with the release of GPT, will be the year consciousness is transformed. GPT-4 is revolutionary because history is consciousness. My late history mentor who influenced the structure of my thinking most16 once cryptically lectured us, nerdy immigrant kids who overinterpreted statistical patterns, that history is not principally about relations of any kind, it’s about the moments when the idea of “we” suddenly transforms. The other recent transformation is of course the one Gerstle fixates on in his book on the neoliberal order, the first https web browsers in 1994. Living in exponential time, 2023 is already a mature phase of this technology, the analogue equivalent of 2009 for web browsers. The rest of the ground may be caught up when GPT-5 and Gemini Ultra are released next year.
The possibilities are endless. The best lecture I listened to this year was Paul Kedrosky, articulating in the clearest way possible what is the economic impact of the massive unmet software demand at small firms filled instantly by a new BLS category 20 million strong of GPT-enabled developers who might have a slightly higher threshold than most but significantly lower than the 80’s and 90’s software priesthood.
To spell it out, the impact is the curve presented by Andrew Ng of Stanford. Before only institutions with billion dollar projects, essentially finance and ad tech two industries which arguably have exhausted productivity gains decades ago, could hire teams of hundreds of software engineers to explicitly write their software code. The threshold gradually declines over time down the curve until you get to small long tail firms which benefit from coding embedded in models which are almost entirely implicit. The productivity gains are much higher in these industries because they have not yet exhausted the low-hanging fruit. As somebody who works on the small firm projects Ng mentions, I wonder just how much these ideas are diffusing into the broader economy in my position on this curve.
The possibilities are also fraught with danger. Will the unthinkingness fundamental to Hayek’s dual identities show his Hyde side, I mused this year after listening to Cowen’s lecture? When LLM’s work with real problems, they are essentially providing the user with some sequence of rules or principles which are embedded in the model.
The distinction is important. Vision products have so much sectoral demand because numerous applications, the legal theorists emphasize, are alien to explicit formulations of rules. If the process is entirely in the background, the user can very easily grow ignorant of the underlying rules/principles. When the underlying principles are trained on valid human judgements, the smiley face, the unthinkingness might not concern. But the cases of drawing from the shoggoth aspects of human judgment are too numerous to mention, and black box of the systems prevents reliable scrutiny. This is the concern of replacing street-level bureaucrats with system-level equivalents. That is not a literature Andreeson and other tech enthusiasts demonstrate awareness of when they glibly talk about replacing government.
I recall a few years ago, by coincidence in the year of GPT-3’s release, my LSE professors telling me that my essay on the potential for ICT-enhanced bureaucracy, Recoding America, to transform government lacked nuance. I was essentially of the view, and remain of the same view, that screen-level is good, and system-level is bad. From those simple propositions, it follows that regulators must establish the AI governance analogue to Maastricht to prevent disruptors like Andreeson from replacing the IRS with 6 AI engineers.17
The principle motivates my thinking about AI generally. There is no doubt that GPT-4 is an extremely primitive tool compared to what will be available in one year, to say nothing of 10 or 50 years hence. And it’s the primitivity that I admire about it. It’s a tool that shows me all that exists in the world, and unlike search engines, compresses all that exists in some digestible format to an output that is directly useful to me. That’s an incredible advancement from this time last year. It doesn’t get everything right which is good! Issues like the jagged frontier and hallucinations prevent me from going on autopilot, and becoming a Hayekian idiot.
In its current form, the learning tool is Lindblom’s preceptoral vision realized. All dimensions of Lindblom’s triad are fruitful, as the relation between student and pupil, AI and users, is also preceptoral. Moreover, note the reversal of classic A/B principle-agent triad under markets and hierarchies, B makes it worthwhile for A or has authority over A. B is the principle in each instance. The burgeoning growth of LLM-driven altruistic collaboration is a fundamental reorganization of human energies, neither a market nor a hierarchy. A does what B wants because A wants to help B. How to ensure AI remains aligned to A, when/if humans are not the only actors in the relation with independent motivation, is the key question.
I want to live in a world where the observations of the researcher remain important, if augmented so we no longer have to worry so much about the frustration of wasted time that comes from solving human-computer principle-agent problems. That desire gets translated as screen-level. Cowen appears to welcome the coming system-level transformation of social sciences. I do not join him. I view it as the latest flavor of End of Theory pronouncements. I think we would be better off reading our street-level predecessors, who might be anachronistic but were definitely wise.
If my Gretchen instincts did not overwhelm, I would be better prepared for the world my professors encouraged me to consider. System-level modes of inquiry are coming, first for business (recall ACI), then the academy, then government, and I erred by simply wishing the anti-human world away. We must start from the assumption that no mast is strong enough to stem the coming tide. The Gyges ring, a device that grants limitless power but renders the wearer invisible, is simply too irresistible.
The Third Movement: A Question of Moral Imagination
On the ninth post of Christmas:
I switched to Chinese to fit into the tweet, but it reads Marx-Engels-Lenin. Promoting his book, The Crisis of Democratic Capitalism, Martin Wolf even went as far to say that all self-respecting social scientists are Marxists. Let’s suspend the political baggage of Marx-Lenin for a moment. Are all self-respecting thinkers in an age of Fourier’s crises plethorique Engelsians?
The key messages of the manifesto which is rarely assigned to college students, Socialism: Utopian and Scientific, resonate more than the one that is. I speculate the former considered the possibility of world beyond classical dynamics, which the 1848 classic mostly did not. The 30 year difference in publication dates could be the reason for Engels’ conceptual advantage. When writing at the beginning of the long 20th century, only sharp observation, not prophecy, is needed to understand the forces set in motion.
By unleashing the Hayekian forces of productive complexity, the abstraction of capitalism is a descent which envelopes all, where the passively governed and actively governing grow ever larger and smaller. Responding to an excellent essay on the Ehrenreich fear of falling concept, I speculated that there are three social classes in advanced capitalist countries.18 Only the working class is governed by wages equal marginal productivity of labor equation, even notionally. The relation is relaxed for the professional managerial class, where hard economicism doesn’t so much apply. They are paid according to their prestige status. The capitalist class exits the relation entirely, and the more they exercise this privilege the more they harden the PMC’s relation to resemble that of the working class. Gas emitted by designed capital inefficiencies, in the language of Michael Jensen, disgorging the free cash flow, transforms liquid into a solid.
The aspect I find perplexing about globe twitter discourse which compares the the relative performance of the first two classes since COVID, and concludes PMC anxieties reflect much-needed rebalancing toward the working class, is the jarring omission. Where does capital enter their picture?
Capital completes the circle, Marchant writes in his history of Big Tech rebellions, “We can all intuit the technological precarity, anxiety, and exploitation…it’s plain to see that informal and algorithmic work is expanding that salaried, benefitted jobs are growing scarce, and if the current trajectory holds, it will engulf us all. Or almost all.” Compare with Engels, writing in between the two tech rebellions, “What Fourier could not indeed see in his time is that the circle is gradually narrowing; that the movement becomes more and more a spiral, and must come to an end, like the movement of the planets, by collision with the center. It is the compelling force of anarchy in the production of society at large that more and more completely turns the great majority into proletarians.”
The tragedy of the collision is their new proletarian status does not in any way accord with their underlying talent, potential energy waiting to be released. In the future Mill foresaw before Engels, we will all be gentlemen:
The division into classes [like the grand classification schemes discussed earlier] has a certain historical justification, it has this only for a given period, only under given social conditions…political domination, the monopoly of culture, the intellectual leadership by a particular class of society has become not only superfluous economically, politically, intellectually a hindrance to development.
Capitalism, disciplining the social imagination to provide whatever form activity disruptive investment happens to identify as valued in the moment, is self-incurred immaturity when applied to people of certain sagacity. The OpenAI debacle revealed that, briefly, they had an island that was not exposed to capitalist discipline. Revealingly, the manner it which the debacle proceeded underscored that even their islands, presumed safe due to their place in the talent distribution, are provisional. The inexorable logic of the collision was in fact inexorable. Sam Altman flatly admitted in his end of year note that fighting the capitalist laws of gravity is foolish.
Socialism in the Engelsian sense instead views the natural evaporating tendency of class, work-leisure, philosophy-scientific distinctions, superficially resembling the primitive, not as something that ought to be instinctively resisted as regression— by artificially reimposing boundaries— but embraced. The sine qua non of development is to encourage sagacity, even if it leads to a temporary decline in the bottom line because they are investing time and resources in activities they do not know how to do, and indeed have never been done before. The good news is primitive small firms, those who do not have the privilege to be exploited by earnings calls, already operate in this way. The future is often seen on the periphery, because the periphery is the motor of history.
With the green transition now in the foreground, there are thousands of activities which must span the globe, often several times over, to find a firm to perform effectively. Unless Engels’ spiraling collision escalates to its tragic apex (recall “pure historicity”), they must all be brought in. I can’t begin to tell you how I’d relocalize my personal operations, but I will tell you that I am experimenting with upgrading forces of production in small ways. Moving from the rules I know to the rules I don’t, the kernel of the idea is small manufacturing outfits might turn into much bigger outfits when widely available 3-D printing merges with AI “god models” to replace the army of quality control engineers who shaped China’s manufacturing juggernaut. I recognize that it is an utterly Utopian thought.
In that process of tinkering, insisting on the unique position of being both above and below the API line, I believe that I am refashioning myself into a capitalist-worker-gentleman. Those porous boundaries were anticipated by Fourier’s utopian thinking. Talent, labor, and capital, how to say which is which? The virtuous path does not concern itself so much with the details of some unrealized Utopia, Engel’s vision of Marxist humanism. So long as I conduct myself with this episteme, focusing on techne to unleash the forces of production, I will be surprised to find that Utopia is firmly at hand.
Tier 1 is upper-level management, tier 2 is lower-level management, and tier 3 is upper-level production. Tier 4 is lower-level production, those sacrificed to Moloch, in Metropolis.
On the tenth post of Christmas:
It’s strange to write in this section, which I planned to reserve for obituaries, about somebody who is still very much alive. But I feel the urge to revise something I said last year about how to select political candidates, which could not have been more wrong in hindsight with how both figures transformed in 2023. Arguing in favor of the preference for the primitive, I wrote Joe Biden and John Fetterman, as second-class intellects, would outperform first-class ones like Jimmy Carter. The horror of the Israel-Gaza war is a startling corrective. In normal times, first-class intellects are uselessly ineffectual. When the world teeters on the edge of chaos, they point to a world that could exist if the second-class intellects were not busy bear hugging with barbarism. The 2020’s like the 1960’s are an Age of Consequences, I told Sahay:
To bemoan the inefficiency of intellect is to miss the point of having intellect. Intellect is, the author of the edge of chaos idea emphasized, to avoid catastrophic sensemaking failures which otherwise envelope all organized social groups. Reflecting on managing Amazon, Bezos perceptively remarked that, “humans are not evolved for truth-telling, but sociality.” The management jargon is institutional isomorphism. As the Rhinoceros virus spreads throughout the social whole, there are certain men and women who read their Niebuhr who stand up and insist, often at great social cost, the current course is insane, “Focus only and solely on the facts and the truth the facts bear out, not that which you wish were believed or would have beneficent social consequences if it were believed.” Moral Man and Immoral Society is to struggle for justice in a sinful world governed by entropy.
Sensing the tumult of the Bush era around him at the University of Chicago, an analogue to the Trump era tumult I’d feel over a decade later at UC San Diego, Tim Sahay, noted the enduring resonance to today. The failures to clean up the “filth” persist, citing his old professor, Bruce Lincoln:
[What] modern discourse treats as impersonal abstractions—things like appetite, anger and falsehood— Zoroastrianism personified as demons. These are the forces which moral people must defend themselves, and these are the forces that tear them apart when their defenses are weak. People who permit their bodies to harbor appetite, anger, and the like find themselves corrupted…Ultimately, they bring death…'Who and what caused his death?’….Mithridates was killed by the vermin who devoured his flesh. ‘And where did they come from?’ Clearly from the stinking filth of his body. ‘And that filth, what caused that?’ His demons, my dear, the terrible demons who permeated his being as a result of his moral failings. ‘What sorts of failings?’ …Probably the lie, the very worst of them all.”
Reflecting on Underground Empire at the LSE,Pettifor noted that colleagues in the developing world, who were a decade or two ago indifferent to the filth, the lie of wetiko, the stench is now overwhelming. Echoing Sankara’s neocolonial call to action in 1987, they exclaim no! The world will not be made in your image, it is you who must change, and acknowledge your global governance rests on the worst kind of filth, a lie.
On the eleventh post of Christmas:
To expose the filth of lies is to reveal the responsibility of the intellectual. Their internal constitution is, struggling with ideas they acknowledge they do not fully understand (i.e. Who are we? Where are we going?), overwhelmed by doubt. Yet doubt is paradoxically paired with historical certainty. I cannot find the tweet, but one of key messages that registered in October is that “When time has past and it will be safe to say what you believe, everybody will say they were against this,” against the backdrop of an utterly decimated Gaza. The right side of history is a cliche, but it also what drives dissidents forward. Speculating on the future of morality from an analysis of a rich dataset on U.S. public opinion, I wrote this year, “resist that instinct which recoils as a Nietzschean slave morality and dare to invent the future of morality, exercise faith, a journey for all.”
Regardless of what you think of his views, John Mearsheimer had a moment last year illustrating the first half of the canvas of what it means to live in a 1956 moment. He was observing the Ukraine crisis from the background and as a result appeared as if he was prophet to those who did not position themselves similarly. The horrors beginning on Oct 7, brewing for a long time —the opening lines of the linked LRB masterpiece begin with apotheosis of predictable savagery—fill in the missing half of what it means to live in a 1956 moment:
The binary treatment of the war in the Western press is mirrored in the Arab world, and in much of the Global South, where the West’s support for Ukraine’s resistance to Russian aggression and its refusal to confront Israel’s aggression against Palestinians under occupation had already provoked accusations of hypocrisy. (These divisions recall the fractures of 1956, when people in the ‘developing world’ sided with Algeria’s struggle for self-determination, while Western countries backed Hungary’s resistance to Soviet invasion.)
Likewise, I would argue that Finklestein, Chomsky, and Khalidi are having a similar moment, viewed as prophets when they simply observed what others viewed as attentionally insignificant. Intellectually curious people with opposite political views feel as if they ought to engage to catch up on what they had missed. Wisdom cannot expect to be achieved through the ignorance of rigid intellectual specialization, many of my co-travelers of polycrisis are increasingly understanding. Prior to Feb 22 or Oct 7, I confess that I had little background knowledge of either, but the Heideggerian concept of prescencing in time, the radical contingency of the historical process, impels a disquiet into thinking.
The key hinge for the dissidents, those who decide to rise up and spark a whole new generation of educated energies which once animated Perry Anderson, again comes to us via Niebuhr, the spiritual discipline against resentment. Resentment is a natural inclination when presented with horror, but we must discipline ourselves from expressing it as that which could potentially frustrate the historical forces set in motion, settling on love. Drawing from Niebuhr, Jimmy Carter’s Sunday School lecture I attended in Plains in 2018 was the parable of Barnabas, to love your enemies. “We look forward to the time when the power of love will replace the love of power. Then will our world know the blessing of peace," William Gladstone.
On the twelfth post of Christmas:
When I set out to do this exercise in 2021, I did not realize the tradition I was summoning. From the Wikipedia entry:
A roman-fleuve (French, literally "river-novel") is an extended sequence of novels of which the whole acts as a commentary for a society or an epoch…The river metaphor implies a steady, broad dynamic lending itself to a perspective. Each volume makes up a complete novel by itself, but the entire cycle exhibits unifying characteristics.
The river, the entire cycle exhibiting a set of unifying characteristics, applies to the course of human development as well as the specific topics I choose to write about in these posts. David Attenborough’s witness statement sets the stage for what I meant when I observed through a mathematics metaphor there are often two ways to solve an optimization problem, one is the serenity of zero, and the other is impossible effort deployed to find the set of natural numbers which will close the equation. He observed of the primitives when he started his BBC career in the 1950’s that they practiced sustainability to their core, “They have been engaging in their patterns of behavior forever, and will continue doing it forever, so it is by definition sustainable.”
David Attenborough's witness statement over the course of one lifetime clarifies the scope of the challenge.
Elizabeth Kolbert, in her abécédaire of the climate crisis, captures the sublime danger of our world-altering might, believing we are masters of everything but commanding nothing. Up until S, Smilian objections, the marvelous testament of homo faber is on full display. Her penultimate entry ends on a deliciously catastrophic parenthetical, the possibility that the late Latour’s Gaia would resist, should we neglect the imperative to apply space age technologies to mimic the primitives but with a natural number set:
If we were to burn through all known fossil-fuel reserves, global temperatures could rise by as much as eleven degrees Celsius, or twenty degrees Fahrenheit (How humanity would keep the oil flowing even as the world drowned and smoldered was a question the researchers did not address.)
Just as there is every inclination for resentment the wise dissident settles on love, Kolbert nonetheless concludes on a bias for hope. It may very well be, like Fermat’s last theorem, that there is no natural number set that can both close the climate system and meet humanity’s insatiable demands, Weber’s unthinking centipede. Yet I believe, as Rosa does when he invokes Weber’s metaphor, that Man is only impelled into the disquiet of thinking when he stumbles, and he can instinctively feel when this occurs. Feeling is as significant as thinking, as we should not think too hard, lest Man fall into Chesterton’s medical mistake, yet another meticulous IPCC report. Sound and fury signifying nothing, Greta Thunberg angrily belted at the UN.
I hope I have demonstrated in these ramblings that we have a breadth of tools that no humans in history have been able to access to correct the course. Those awesome space-age tools, Berman intuited, might be like magnificent palaces that humanity at scale was not meant to live in, the social crisis of acceleration Rosa diagnoses.
The reference that comes to mind when thinking about potential transformations to organizational structures of different types, which I should revisit more carefully soon due to this new dimension, is Mary Douglas’ How Institutions Think (1986).
I’m sure the OpenAI people would go to great pains to emphasize that GPT-4 is no code monkey, though it is not lost on people that this is currently the overwhelming use for the service. Oh and to make good “more extreme” memes.
I’m very drawn to theories of the impoverished sophisticate, that sometimes man is mismeasured in ways that don’t accord with his environment in a manner that drives development, because it aligns with my theory of history as guided in the paradox of development. I think history is paradoxically often made far from most advanced center, whether it be the social basis which enabled the big bang corruption reforms and the Nordic social welfare model or the scientific expertise which gave birth to modern China under Deng’s refashioning of the guerrilla legacy in the service of the 4 modernizations.
It’s impossible to read Wansleblen’s The Rise of Central Banks or Tooze’s Crashed without marveling at the creativity embedded within the escalating adaptive strategies to maintain the high-wire balancing act. The Wansleblen book and his public commentary in particular emphasize the “everything is forever until it is no more” theme suggesting that there has to be a way out into an alternative form of creativity. And yet, Wansleblen flatly acknowledges that his book falls squarely within Chesterton’s medical mistake critique. Purely rational descriptions drain the will to act.
I don’t have the patience of mental masochists and lovers of obscure shit to read Heidegger directly but the concept of presencing in time, impelling a disquiet into thinking, points in the direction of the psychological state I am trying to summon by the word “qualia”. Like Heidegger, I share a problem of using language to communicate.
This is unfair to Summers. The real 9-ft aliens are the tech billionaires presiding over neofeudal settlement, Zuck, Andreeson, Musk, Bezos, where they say serve you, but really they are manipulating you to serve to others for their consumption. You’ve Been Played.
Presumably the humanist creator of Star Trek, Gene Roddenberry,was referencing John Keats, “Beauty is truth/ truth beauty/ —that is all/Ye know on earth,/and all ye need to know.”
As argued in a superb book, Blood in the Machine: The Origins of the Rebellion Against Big Tech. The contrast of course is on who bears the stress as a result of the shocks. With the exception of Britain, nowhere in Europe at the beginning of the 19th century was particularly urbanized with many struggling to meet basic needs. Even here, the echo rings through. Let the developed core stand in for Britain and the developing world stand in for continental Europe and you might better understand that most of the world is preparing to subsist on a diet of potatoes, made worse by lack of access to debt financing.
There is an interesting reference dropped to me from PolicyTensor, The Analysis of Social Change, which has an extended explanation that the reader might misunderstand our use of “religion” abstracted beyond particular time and place in an anthropological sense. Philosophical needs and religious needs are the same, the Wilsons write.
The hodgepodge collection of bits here, for better or worse, is completely owed to him. He struggled with high school kids who took the path of least resistance and shared their “feelings” about history. No, your feelings are not interesting; I want historical evidence, and only protracted struggle between pieces of evidence. He surely intended this structure to be simply a ladder that is kicked away at higher levels of study because Collingwoodian idea of history of personal interpretation is vital, but was wise to intuit that 16 year old students had not yet struggled enough to wield that responsibility. I’m still in the process of kicking away the ladder.
