Of an average, inconspicuous human being.

Like a lot of pontificaters, professional and amateur alike, I’ve had recent AI developments on my mind. Scary stuff, in many ways, all depending on the sort of newsletters you subscribe to. I was also reading Vasily Grossman’s Life and Fate, the great Russian novel of World War 2, finished in 1960 and very far from science fiction. One chapter, short enough I can type it all out, jumped out at me for reasons that should be clear enough.

An electronic machine can carry out mathematical calculations, remember historical facts, play chess and translate books from one language to another. It is able to solve mathematical problems more quickly than man and its memory is faultless. Is there any limit to progress, to its ability to create machines in the image and likeness of man? It seems that the answer is no.

It is not impossible to imagine the machine of future ages and millennia. It will be able to listen to music and appreciate art; it will even be able to compose melodies, paint pictures and write poems. Is there a limit to its perfection? Can it be compared to man? Will it surpass him?

Childhood memories … tears of happiness … the bitterness of parting … love of freedom … feelings of pity for a sick puppy … nervousness … a mother’s tenderness … thoughts of death … sadness … friendship … love of the weak … sudden hope … a fortunate guess … melancholy … unreasoning joy … sudden embarrassment …

The machine will be able to recreate all of this! But the surface of the whole earth will be too small to accommodate this machine — this machine whose dimensions and weight will continually increase as it attempts to reproduce the peculiarities of mind and soul of an average, inconspicuous human being.

Fascism annihilated tens of millions of people.

Vasily Grossman, Life and Fate, chapter 51. Translation by Robert Chandler

June 21, 2023

This title will not subsidize another.

An interesting column from Dan Sinykin in the Times:

A career like Mr. McCarthy’s, with its long gestation before a blockbuster second act, would be nearly impossible to repeat now. An author without an agent or a track record of book sales would never gain a hearing at a major publishing house. And a state-school dropout in his early 30s would face slim odds of becoming a prizewinning author, as an M.F.A. from a prestigious writing program has often become the price of entry for splashy literary fiction debuts.

It’s impossible to know what kind of writer Mr. McCarthy would have developed into without decades in which to hone his singular voice. But contemporary success stories about novelists tend to have a very different aspect: They’re stories like that of Colson Whitehead, who followed up his well-received 1999 debut, “The Intuitionist” with novels that deftly navigated genres before reaching a new plateau by winning a Pulitzer Prize in 2017 for “The Underground Railroad.” Or Bonnie Garmus, the author of “Lessons in Chemistry,” whose very first book became a runaway hit in 2022.

Cormac McCarthy Had a Remarkable Literary Career. It Could Never Happen Now.

I’m very interested in how the publishing industry conditions shape the literature that is produced, but I feel like there is more to the story than corporate consolidation. The nature of the reading public has also changed. The generation of young men in the trenches of World War 2 who eagerly read the first mass produced paperback novels are gone. I was reading reporting from the Ukrainian front line only a few weeks ago that featured a young soldier on the front line posting to social media. All creative conditions are transient and the only constant is that artists are forever waiting for favorable winds to arrive.

June 20, 2023

Any of you writing a book?

I’m deep into The Dud Avocado now. We run into the following amusing incidental character on page 170. (All you really need to know if that Somerset Maugham was an English novelist and national celebrity. His most famous work, Of Human Bondage, is strongly autobiographical).

Boy, we met a real nut on the beach today. A skinny young American with a fierce black mustache called Hugo McCarthy. He’d just left the South of France a couple of days ago, where he’d been kicked out of Somerset Maugham’s villa in Cap Ferrat for the third time. Every time he thought about it he became absolutely doubled up with rage, his whole body trembling under the impact of his emotions.

“Where does he come off giving me the bum’s rush?” he steamed. “Who the hell does he think he is anyway? He needs me, boy, I don’t need him! I’m the colourful eccentric all these characters write about. Hey, do you know I’m in three books already? Met an Englishman called Tynan in Spain a while ago and he put me in his bullfight book, and a couple of Americans I ran into last year skiing at Klosters — wouldn’t be fair to tell you their names until I hear how it comes out, but they’re very well known — anyway they’re still fighting about which one’s going to have me in his.”

The Dud Avocado, Elaine Dundy

It is no great observation that many famous literary creations are holograms of real people. I think of Ignatius Jacques Reilly and Malachi “Buck” Mulligan, just off the top of my head. I’m reading The Dud Avocado itself as being pretty autobiographical, so our rascal Hugo McCarthy might likely himself be an amusing echo of an actual young man. I certainly don’t want to suggest Dundy wouldn’t be capable of such invention, but people of McCarthy’s cut can certainly be found today among the crowd who tailor their personae to winning a coveted spot on some reality TV show. The producers of such shows, I understand, are entirely willing to engage with such a market, while Somerset Maugham, it seems, was much less so.

June 19, 2023

…and a triptych of 3-D dog posters…

Robert Gottlieb, the editor of a sizable and influential portion of American literature, passed away the past week. It feels like there has recently been a steady stream of literary, cultural, political figures passing away in their nineties. I remember when I first came across the phrase “the pig in the python” reading Douglas Coupland’s novel Generation X. It describes the sheer demographic bulk of the generation born post World War 2. But Gottlieb, born in 1932, was before even that. He was young enough to avoid the war, but of age to provide to the actual baby boomers when they were ready to read. We haven’t really got to the real pork yet.

I was taken with a certain detail in the Times’ obit.

A sickly, lonely, unhappy child, he sought refuge in books, which he sped through. As a teenager, he said, he read Tolstoy’s “War and Peace” in a day and Marcel Proust’s monumental “Remembrance of Things Past” in a week.

“I would read three to four books a day after school, and could read for 16 hours at a time,” he told The Times in 1980. “Mind you, that’s all I did. I belonged to three lending libraries and the public library.”

https://www.nytimes.com/2023/06/14/books/robert-gottlieb-dead.html

I’m not sure it is possible to really read War and Peace in a day. Maybe there is quite a lot of truth to the story, but it feels akin to the fact that Montaigne was supposedly raised to speak Latin as his first language in 16th Century France. I don’t want to doubt the precocity, but such exploits usually have a few caveats.

To whatever extent that it is true, I am nevertheless jealous. To sit and read in such relentless fashion I need to either be on a flight, or somehow similarly trapped in place. There is some immersive quality unique to literature when you force yourself onward into a novel, beyond the easy limit of your attention.

June 18, 2023

The Dud Thinkpiece

The New York Times has profiled some young people who briefly ran a tiny circulation newspaper in New York during the pandemic and are now launching a brand-spanking-new online publication, The Byline. Running a small press newspaper runs the obvious danger of being an exercise in nostalgia, but now that danger extends to running upstart an upstart, online magazine.

Ms. Lorenz, a former New York Times reporter who is writing a column for Byline, roll-called a number of defunct or diminished sites: “There was this middle ground when I was coming up — xoJane, Jezebel, Rookie Mag, Hello Giggles, Man Repeller — these midlevel media outlets where you could get your feet wet.”

“It’s all gone,” she said, “and no one has built anything to replace it.”

I dipped into one article about the Wikipedia that was getting linked in places, and found it strange. I won’t say it was bad, but the first thing the writing reminded me of was chat-gpt. (“Their tireless efforts are the building blocks upon which the internet’s foundations rest, with Wikipedia standing tall as a testament to their ingenuity and collective wisdom.”)

As it happens, I’m reading The Dud Avocado, which features pretentious Americans in Paris, self-conscious and conspicuous in their efforts to make a literary scene happen.

“Listen, Jim, about this new magazine we’re getting under way: first of all I want to make it quite clear, see, that we’re not having anything to do with all this effete chi-chi they’re trying to unload on us now.” A would-be Editor disdainfully brandished two small Paris-American magazines he’d just been sold. “I mean ours is going to be really experimental, in the true sense of the word, for Christ’s sake. These other bastards must be walking in their sleep or something. Look at this-” he flipped through the pages scornfully. “A reprint of an early Spender poem and a lousy Ugo Betti translation — Ugo Betti, for crying out loud.”

The Dud Avocado, Elaine Dundy

Maybe I should pitch my own lousy Ugo Betti translation.

June 17, 2023

I, for one, feel like the machines have learned enough.

Against all possible odds, I read a poem in the New Yorker and actually liked it. Fishing in the Blood by Donika Kelly. It is a poem that tells a story without an excess of obliqueness, obscurity, or vagueness. A fragment of dialogue that takes a little effort to parse, but grows on you as you do parse it and get to know the speaker. It reminds me a lot of Toni Morrison’s first novel, The Bluest Eye, to the point that it almost feels too derivative. I wonder to what extent anything new that feels properly literary will be leaning on giants of 20th century literature. The obvious rejoinder is that some version of this has always been the case.

In any case, it feels good to have found something that you would never have served up to you by one of the many algorithms that haunt our world. A fucking poetry editor made the decision to print it. And flouting all good judgement I decided that I was going to try reading one of the poems that week.

June 16, 2023

Mirrorflaking the ocean’s steely surface

After two decades of going over-schedule and over-budget, the James Webb Space Telescope is finally proving that there is no such thing as over-exposure when photographing deep space. Arriving at its final destination in orbit about the Sun, blossoming and unfolding its hexagonal petals, the JWST began soaking up photons — photons coming to the end of a long journey from the very distant past. Aside from some spacey looking, Chris Foss-esque, snapshots, serious technical data is being collected. By carefully observing the flickers of stars, scientists are measuring light spectra as distant planets cross pass over the face of their sun. From these spectra, the composition of elements on those planets can be deduced.

Most of us will never get to see the Earth from orbit or walk on the moon. The solar system and beyond will remain bright dots in the night sky. The craft of Astronomy is to deduce facts from a limited vantage point, with inevitably limited means, about phenomena that you will never observe directly. Astronomers refer to the chain of reasoning that they have employed as “climbing the celestial ladder”. It’s the oldest scientific game going, with shockingly accurate estimates of the earth’s size attributed to Eratosthenes in the third century BC.

We differ in many respects to those Greeks, not least of all in the extent and power of our technology. But here is another: we possess powerful convictions about the future, animated at least in part by science fiction. Some, when the JWST data is all brought together and processed, will be more than merely curious about the theoretical existence of life “out there”. Those certain people will be looking at those worlds as potential destinations for the future of our species.

When Herge had Tintin set his ligne claire foot upon the moon before Niel Armstrong, so he had to imagine how it could happen. There was a literature at the time, written by scientists and engineers, on the subject of rocketry and a future moon expedition and Herge consulted it. The aim was realism and in that Herge was remarkably successful. The field of “hard SF” — science fiction which sincerely aims to pays its dues to the equations — has shown much less restraint, but nevertheless seems to harbor the belief that they are anticipating similar future developments. Many authors and readers believe these stories have social utility in either inspiring us toward grand feats or in anticipating the inevitable challenges that await us. Either way these writers become prophets, sensitive to the whispers of the goddess Science in their ears.

This review of Aurora, by Kim Stanley Robinson, will stray deeply into the “spoilery” territory some of you like to avoid. Territory where I’m not so much assuming you have read the book, but that you are fine with me divulging the key contours of the plot so I can explain my grand theory about what is really going on. For my part, I started reading Aurora with a good idea what lay ahead. The story, as far as I understood it going in, was going to be a fictionalized illustration of a guest essay Robinson wrote for Boing Boing many years ago — that I read at the time — which laid down a thorough account of the reasons why we aren’t going to be colonising other planets or doing much of any travel outside our solar system.

The premise of Aurora is that humanity reaches for the stars anyway. We alight upon a generation star ship as it reaches the end of a 160 year journey to Tau Ceti — a nearby star possessing Earth-like. The generation star ship is a tool for science fiction writers; a means of carrying mankind across the unfathomable distances between stars. Rather than hyperdrive or cryosleep, you imagine a spaceship as a kind of grand terrarium in which a living, breathing, teeming, reproducing lump of Earth’s ecosystem, humans included, are sent off in a self sustaining bubble. A nuclear reactor provides the constant input of energy required to avoid entropy, and with the turning of a few pages hundreds of years pass, generations rise and fall, and we arrive at their destination.

There are characters, but only three are really important; only two are human. Devi, the ship’s chief engineer, whose existence is devoted to fixing the alarming array of problems that are presenting themselves, and Freya, her daughter, struggling to follow in her mothers footsteps despite lacking inclination toward anything mathematical. Then there is the ship’s AI, who might even dispute their status as a character at all, but has been cultivated by Devi into a storyteller who narrates most of (and I would argue the entire) novel.

Readers responded most strongly to the AI narrator, won over by the ship wrestling with the way we employ metaphor and simile in our language. But this amounts to the ship transitioning from writing with the dispassionate, exhaustive, technical register you’d expect from a computer, to the kind of cliched idiom you’d encounter in the prose of an eager yet unimaginative writer. That is until there is some geography to be described, in which case they invariably write like the singular Kim Stanley Robinson.

To my own liking, Devi is the most compelling character. She lives with a constant sense of anger. Not at anyone who shares the ship with her, but at those predecessors who volunteered themselves and condemned their offspring to the journey. As the chief engineer, she is responsible for addressing the oversights and shortcomings of the original design. As a consequence, she understands better and earlier than anyone else on board how truly nightmarish their reality is.

Kurt Vonnegut’s sixth of eight rules for writing seems especially relevant here:

Be a sadist. No matter how sweet and innocent your leading characters, make awful things happen to them—in order that the reader may see what they are made of.

In this respect Robinson uses the generation star ship as a terrible predicament for our characters. They aren’t intrepid voyagers; they are poor souls born in a tin can hurtling through the void. By the end it is the great, great, great, great grandchildren of those who first set off from earth who arrive at the destination. One of the all time great sci-fi speculations, but awful for anyone who would be caught, trapped, sealed inside such a thing.

It’s not Robinson who is the sadist — it’s all of us. The trick in storytelling is allowing the reader implicate themselves in the actions and events of a story. In Aurora’s case we are implicated as science-fiction readers, eagerly anticipating mankind setting foot on another world, with little consideration of the actual fate of the generations of voyagers stuck on the starship. We are with that first generation of volunteers who effectively consigned their descendants to their bleak and unpromising fate.

We are to learn the hard way that “habitable” and “Earth like” are terms-of-art for astronomers, applied to a broad family of planets. What resembles Earth from a great distance might not be a place we would want to live. Aurora is the planet the colonists settle on with its limited prospects: oceans, wind, and oxygen. Filled with relief, they rush off the ship that has been their inter-generational prison and start building a new life. Unfortunately, those who expose themselves to the ostensibly breathable atmosphere start having extreme adverse reactions, quite likely from some mysterious microscopic alien organism that they can barely identify. Thus, they find themselves caught in another great literary trick: the double bind. Either they choose to settle an Earth-like planet can potentially support life, in which case it will be almost certainly poisonous, or they choose an Earth-like planet that is inert and unlivable, on the promise that terra-forming will make it livable in a few hundred years.

It is a long way to have traveled to arrive at such a truth. To be betrayed. The population struggles to accept this reality, falling into violent civil war, and ultimately having to divide itself. The ship is split in two. Half of them remain, believing that colonization will still be possible, while the other half set off on a return journey to Earth. At this point, quite conveniently, our heroes are presented with newly invented hibernation technology that will allow them to sleep through the return journey. Not only sparing them from famine as the internal balance of their limited ecosystem begins to collapse, but also to allow those who arrived at Tau Ceti to confront the society that sent them in the first place.

The final chapter of the book has no clear narrator (although I believe that there is one implied). With insufficient means of slowing it’s approach, the ship performs a sequence of reverse gravity slingshot manoeuvres to slow their trajectory sufficiently so that the remaining colonists can be ejected back to earth, while the ship itself is lost into the sun. Freya and her shipmates arrive on earth, discovering in it a to be a home like nowhere in the universe. Perversely, many on earth are upset that they returned, feeling that they betrayed mankind’s true destiny. It is on this final ironic note that the novel might have ended, but instead we finish with an extended passage where Freya goes to the beach and enjoys the waves. This final passage has proven particularly dissatisfying and mystifying to many readers.

With the ship perishing in its final orbit around the sun, the reader might be obliged to conclude that the ship couldn’t possibly be narrating the final chapter. That is to forget why the ship was narrating at all. The ship’s narration was an exercise set by Devi as a means of explaining to the occupants on board the ship the nature of their predicament.

Devi: Ship! I said make it a narrative. Make an account. Tell the story.
Ship: Trying.
[…]
Devi: […] Do what you can. Quit with the backstory, concentrate on what’s happening now. Pick one of us to follow, maybe. To organize your account.
Ship: Pick Freya?
Devi: … Sure. She’s as good as anyone, I guess. And while you’re at it, keep running searches. Check out narratology maybe. Read some novels and see how they do it. See if you can work up a narratizing algorithm. Use your recursive programming, and the Bayesian analytic engine I installed in you.

While we can assume that everything up to a certain point is true, I don’t see any reason why the ship wouldn’t extrapolate further, simply because each story needs an ending. What better way to explain the situation than to imagine what might likely happen? Allow them to foresee a future as it might likely be. The ship would be doing what Robinson argues science fiction writers have failed to do. That the narrative was in large part an AI hallucination would explain many things, not least of all the convenience of the hibernation technology. But also the final passage on the beach, which closely tracks the opening passage of the book, set out on a lake in one of the ship’s internal ecological containers. The ship is solving the “halting problem”, frequently alluded to, with a recursive, oroborean structure, starting where we began.

Back in late 2020, and early 2021 I read a very different generation star ship novel (or really a quadrilogy of novels). The Book of the Long Sun by Gene Wolfe. As specific as Aurora’s preoccupations are to Robinson, he readily alludes to it’s influence:

One of the best novels in the history of world literature, Gene Wolfe’s Book of the Long Sun and Book of the Short Sun, a seven-volume saga telling the story of a starship voyage and the inhabitation of a new planetary system, finesses all these problems in ways that allow huge enjoyment of the story it tells. The novel justifies the entertaining of the idea, no doubt about it.

The central conceit of Wolfe’s story is that the population of the Whorl have forgotten that the live in generation star ship — something that Robinson’s characters could only wish to forget. Like all of Wolfe’s works there is a literary preoccupation with the internal logic and the implications of who is telling the story. In the case of the Long Sun, save a dropped pronoun, up until the very end the reader must assume that the text is narrated by the omniscient third person.

I don’t think that Wolfe saw himself as a practitioner of hard SF, and although he had given thought to how the Whorl worked, it all seems to be in service of providing the reader with a fantastic literary vista — a setting that forces the reader to reorient themselves in reading the text. The real appeal of SF, as far as I judge it, is not that the authority of science grants the writer powers of prognostication, but rather a certain means of suspending disbelief, and then remarkable destinations to carry the reader away off too once their disbelief is in flight. It has been glibly asserted that SF needs to be more than jet-packs and ray guns, but Wolfe himself, even in his literary shenanigans was all about the jet-packs and ray guns. The problem was never the paraphernalia. The problem with a lot of SF was that the writing was not very good.

The ship was pulling its punches. I half anticipated the ship to arrive back at Earth to find that reintegration of the passengers into Earth’s ecosystem was impossible. How many new viruses would have evolved on Earth in their absence? How dangerous would their own bacteria have become to humans on Earth? Or perhaps they would be received back on Earth with abject horror. Inured to their own true state, they don’t appreciate how irradiated, starved, inbred, and traumatized they. But these would have been endings that would have not have suited the ship’s purposes, nor Robinson’s inclinations as a writer.

In the end mankind’s colonisation of the galaxy resembles less of a science based prediction of the future, than a kind of secular heaven. It lives, like heaven, in our imaginations, where it can be left unchallenged, save for when we encounter true believers. The aspirations of Star Trek, like many religious traditions, are better inherited in a flexible, non-literal form. But that also has grim possibilities as a divide opens up between the naive believers in our science fiction future and the better schooled theologians who will inevitably have to choose their words carefully so as not to alienate the faithful. In the case of the novel, the readers are left to imagine how the passengers on the ship in Aurora might have eventually reacted to story they have presented to them. Back in the real world a work like Aurora entering the “canon” of SF might inoculate us, but if the resistance to accepting Joanna Russ’s We Who Are About To… into any kind of canon is any indication, Robinson has his work cut out.

August 28, 2022

The Long Tail Twitches

The 20th Century promised us flying cars, jet-packs, and Mars colonies. And how could it not deliver? It was already happening. The atomic age arrived, ending the Second World War in terrifying fashion, nuclear power was ushered in, and men walked on the moon. But where exactly did we fall short in achieving the envisioned grandeur of tomorrow? Was anything less than Star Trek going to be a disappointment?

The 21st Century birthed a fresh litter of new promises and speculation; now it was the internet that was going to transform mankind. And how could it not deliver? The internet was already here. We were connecting and reconnecting via social networks, being blessed with the unlimited email inbox space, discovering with instantaneous search, guiding ourselves through meatspace via smartphone GPS, and abandoning TV for streaming. You have to separate the broken promises, and from those which became a nightmare.

One promise was of particular interest for those who, like me, feel some investment in our arts and culture (in my own case, being somewhat ‘bookish’). It was argued in a Wired essay, then spun into a book, that the internet would fundamentally change the market for culture; supply would meet demand more responsively and more dynamically. We had unmet needs stuck out in the “long tail”. The niche and otherwise obscure would reach the audiences who would appreciate it. In many ways it has. But the true implication of the long tail is that most cultural output is now the krill that the big conglomerate whales feed off. No individual artist living in the long tail will make any real money, but a corporation taking a 10% cut from the entire the long tail has a healthy return. We won’t be leaving the old gatekeepers behind any time soon and we still live under the tyranny of the lowest common denominator.

The new media landscape that has emerged has deeply frustrating shortcomings — even for increasingly well served readers. As Matthew Claxton observed in his newsletter on the phenomenon of the Brandon Sanderson kickstarter, if you go to Sanderson’s Amazon webpage for any of his books, the algorithmically generated recommendations put you at the bottom of a pit full of other Sanderson titles. The algorithm sends you to books you are likely to buy, which ultimately means bestsellers very similar to the bestseller you are already looking at. There is no effort to introduce you to new releases you might be interested in, or take you anywhere off the beaten path. Amazon has no interest in doing the things that bookseller enjoy doing, which must be utterly infuriating for booksellers who are learning the hard way just how irrelevant they are to the business of simply selling as many books as possible. Amazon could have been interesting if they weren’t so singularly focused on their algorithms.

What kind of bookseller would feel any pride in just putting the most obvious bestseller into the customer’s hands? One who only cared about making money, probably. One who had no investment in the treasure of discovery, certainly.

Let’s face it though: all of this talk is tiresome. The failures of our futurists is far less intriguing than the question of how we actually stumble upon our treasures of discovery. There is no reading list for life, and it is difficult to even talk seriously of a canon. It is more like we have inherited what is both a legacy and still growing, thriving enterprise. Both past and future works outpace us as more gets published then ever, and older works are rediscovered and re-evaluated. Every book listicle is a desperate attempt at navigating this reality. Engagement with this leviathan is no small feat: a book takes an unusual degree of engagement, in terms of time, concentration and intellectual commitment compared to most other narrative art forms.

To look back and wonder why I read any particular book is to confront free will itself. Something that the ideology of big tech pays some tribute to in all their talk of providing us with more “choice”. But it is reassuring to look back and find half a mystery and plenty of serendipity.

Some books I have read in large part because they were sat on a shelf before me, especially when I was younger. I read Thomas Harris’ counterfactual thriller that I found in the Welsh holiday cottage my family stayed at on the strength of the implied horror of blurb alone (what if…?). A similar story with Witches Abroad by Terry Pratchett, with the Josh Kirby cover art that drawing me in rather than the blurb. It was bound to happen sooner or later — there are enough copies of both Harris and Pratchett out there that maybe it was only a matter of time.

Sometimes it is someone else’s initiative that leads me over the threshold of the first chapter — I make a point of reading the books that friends and acquaintances recommend to me. HHhH by Laurent Binet. Rant by Chuck Palahniuk. Death at Intervals by Jose Saramago. Me Speak Pretty One Day by David Sideris. Cuckoo’s Egg by Cliff Stoll. Fifth Business by Robertson Davies. 4321 by Paul Auster. A Walk in the Woods by Bill Bryson. Recommendations removed from book marketing and publishing log-rolling, and read free form the usual burden of expectations. I really can’t think of any that have been remotely dissatisfying.

Then there are the recommendations of famous and celebrated writers. This is treacherous ground, with all the log-rolling, blurbing, favors, politics, and insufferably chummy niceness that ultimately muddies the waters. You rarely hear an author expressing dislike for another authors work, but any serious writer must have stronger opinions than they let on. A reader with a developed sense of taste should be discarding books they find weak or disagreeable, and I find it hard to believe a writer the stones to make strong creative decisions if they can’t abandon a dismal novel.

I can think of two writers whose recommendations I would trust. This isn’t necessarily correlated to my opinion of them as writers, just on how their recommendations seem to have worked out. 1. Neil Gaimain — who seems to have a consistent knack for championing and recommending great work in sci-fi, fantasy, and horror. He led me to Gene Wolfe, Susanna Clarke, and Jonathan Carroll. 2. Will Self — who, when he isn’t musing on the future of the novel, recommends some seriously interesting reads — The Strange Last Voyage of Donald Crowhurst by by Nicholas Tomalin and Ron Hall, and My Father and Myself by JR Ackerly. (I also need to get around to Riddly Walker by Russell Hoban, which he has been effusive about.)

It is worth noting that The Book of the New Sun by Gene Wolfe has a more complicated provenance than the Niel Gaiman recommendation. I have a vivid memory of being in high school, and reading a glossy paged coffee table book that must have been Science Fiction: The Illustrated Encyclopedia by John Clute. I pored longingly over the reproduced cover art and reading accompanying text. One book sounded particularly arresting. A singular reading experience where the reader is introduced to what is ostensibly a fantasy setting before it is gradually revealed through various clues and hints, that it is fact the Earth of the distant future as the Sun is dying. It was a quadrilogy, which back then seemed like an unthinkable commitment, even if I could track down a copy. Yet the memory of the book lodged itself deeply in my mind as something that I should read. Yet I didn’t memorize the author or title, so I really have no idea how I managed to rediscover it later.

And then there are the lists. There are so many book lists, and you might wonder what possible use they may have. One list in particular — 1001 Books to Read Before You Die — published in book form, left a particular impression on me because I quite seriously set about trying to read them all. I was young and the attempt could not have lasted more than a year. But this would have been how I discovered Philip Roth, David Foster Wallace, Ian McEwan, and many other “obvious” authors. I only got about a third of the way through A Suitable Boy by Vikram Seth, and I should probably get around to giving it another go. There was something incredibly empowering about just going in blind, irrespective of any perceived “difficulty” the book might have. I suspect it liberated me as a reader.

There have been other lists that have been useful. The New York Times Best of the Year list and the Booker prize shortlists have been worthwhile. (A Tale for the Time Being, bu Ruth Ozeki being a stand out discovery). Sometimes books just have reputations: 1984, A Brave New World, The Selfish Gene, and Sapiens. Sometimes they got mentioned repeatedly on the New York Times Book Review podcast, and the authors interview very well — so I had to pick up Educated by Tara Westover and Priestdaddy by Patricia Lockwood. Say what you like about the NYTBR, but having little interaction with the humanities side of academia, the highly networked worlds of genre and lit-fic, nor publishing more generally, they have been the best means of access available to me for the vicarious thrill of insider book-talk.

And who wrote those lists? Critics, you’d hope. Book reviews do in fact occasionally drive me toward reading a specific something. It was newspaper reviews that led me to Nothing Is True and Everything Is Possible by Peter Pomerantsev and Countdown to Zero Day by Kim Zetter. I’ve tried a few books on the strength of them getting an A on The Complete Review. And was given the impetus to read Aurora by a New Yorker profile of Kim Stanley Robinson from the past year. But disappointingly it seems like it is the general critical consensus that moves me to read something that any one given critic. I guess James Wood is responsible for bringing Karl Ove Knausgård and Elena Ferrante to our attention, and I have read books by both of them. But it wasn’t James Wood’s writing that sent me to the bookstore.

Unfortunately it seems that my need to read a book accumulates within me as mysteriously as the workings of any algorithm. I badly needed to get around to reading The Secret History by Donna Tartt and The Sparrow by Mary Doria Russell, but I couldn’t point to any one particular recommendation. And I read Gilead by Marilynne Robinson understanding the general critical acclaim it holds. It’s not unusual for me to read a book review and conclude that the book in question sure sounds great, only to file a mental note away in the overflowing mental filing cabinet.

And then there is the fact that one book can lead to another. I’ve spent a fair amount of time on various author’s Wikipedia pages trying to make sense of their influences. I’ve never anything like a systematic study, but I read Vance’s Tales of a Dying Earth because of Gene Wolfe, or and dipped into Dickens because of how much his novels have meant to Donna Tartt, among others. Authors may bridle at the question of where they get their ideas from, but tracking down their influences is fantastically easy. I could of course go about it the other way around and enroll in a “Great Books” course so I could convince myself that there is a coherent trajectory in literature that arrives at the present, but I think that’s better left as a vague aspiration.

A great deal of ink has been spilled in service of justifying extensive personal libraries of books that couldn’t possibly all be read in the owners lifetime. But their principal justification must be the ready richness of possibility that lies open before you when choosing the next book. We aren’t whales consuming books like krill. A book takes up residence in your inner life for days, weeks, months, or years. The cerebral furniture has to be arranged to accommodate the guest. As readers it doesn’t make sense to think about the “long tail” when the kinds of book I’m reading lived out there long before the internet came along. The long tail isn’t even a particularly meaningful concept then to readers like me except perhaps in reference to the way certain books can stay with us long after we’ve closed their covers.

August 4, 2022

The vibe moves

In 1633 Galileo was put on trial by the Roman Catholic Inquisition. The ailing mathematician suffered through the indignity and ordeal of the trial to be declared guilty of advocating heliocentrism — the theory that the sun, not the earth lay at the center of the universe. Galileo’s book, “Dialogue Concerning the Two Chief World Systems”, was the principal piece of evidence presented against him. But as Dan Hofstadter repeatedly belabors in his book The Earth Moves, there was no material debate on the merits of Galileo’s arguments. The only relevant question was whether Galileo had usurped the authority Catholic Church over interpretation of scripture.

So the Counter-Reformation was fighting a losing battle against a freer, and often less literal reading of the Scriptures. Yet if there was one thing that had concerned the Council of Trent, it was the possibility that laymen would decide for themselves what passages in the Bible could be interpreted other than literally. In fact, the issue of the earth traveling about the sun had little if any bearing on the Catholic faith. But the notion that persons without theological training could decide for themselves to read this or that biblical passage in a non-literal sense constituted a mortal danger for Catholicism in the early seventeenth century.

Hofstadter’s book focuses on the trial, while also giving background on Galileo’s observations of the night sky through his recently invented telescope. We are not treated to a full biography of Galileo, nor the kind of exposition — which I would have certainly benefited from — on the nature of the reformation, counter-reformation, or the Roman Catholic Church of the time . Hofstadter’s own interests clearly lay in discursive discussions about Galileo’s engagement with the art and culture of the period.

There is much that is interesting, but also a great deal that is frustrating. After outlining the events of the trial, Hofstadter hedges against saying anything too definitive about the affair, conceding that both the event of the trial, along with its final verdict, may likely have been the consequence of unknown Papal intrigues and the obscure politics of the time. The book doesn’t seem to know what to do with certain pieces of context. Take the following insight into what we know of an individual’s religious conviction:

We do not know what Galileo or anybody really believed at this period, since religious belief was prescribed by an autocracy and heresy was an actionable offense. If one had misgivings, one kept them to oneself, so it would be naive to take religious ruminations penned in the papal realm or its client territories at face value. The Inquisition’s own records confirm that many people harbored reservations and heretical beliefs: before the Counter-Reformation, they had been much more candid about them.

Galileo had built a telescope that provided a tiny, limited window that gave him a better view upon the solar system. This insight shaped a scientific conviction that led him to stray on territory that the Church had claimed authority over. But it didn’t have to be science. If anything, science was at the fringes of what the Church controlled. Moral, social, and religious matters were the principal victims. Science just happened to have been the issue that the Church managed to look definitively foolish over. Frankly their other strictures look similarly bad today, at least to my eyes. The extract I just gave above seems to hit on something at least as important as the actual science that Galileo practiced. It almost seems to serve the Church to frame the affair as “science vs religion”, rather than as “religion vs freedom of thought”.

Hofstadter clearly loves Galileo as the Renaissance man, immersed in the art and culture of his day. I sensed a real nostalgia for the pre-Two Cultures world. There is a valor and a virtue that is popularly recognized in those early scientists — the “genius” we ascribe to those who were the first to figure certain things out. I am perhaps sensitive to a certain kind of sleight made against the “institutionalized” scientists for today who are fantastically empowered by the inherited work of earlier scientists. So great is the inheritance that it inevitably dwarfs any possible contribution they can make. It is a sentiment often derived from the valorization of those heroes of the scientific revolution. Sure, you can hear them sneer, you’re clever, but you aren’t a genius. I don’t sense any of that from Hofstadter. I could imagine him saying something more like: Sure, Galileo’s reasoning reads as scientifically illiterate to us today, but unlike academics today, he could actually write.

In case you don’t believe me about how Galileo’s reasoning reads today:

Aristotle had had no conception of impetus, and thus no conception of motion corresponding to what we may see and measure. He thought that the medium through which objects travel sustains their motion. By contrast, Galileo wrote “I seem to have observed that physical bodies have physical inclination to some motion,” which he then described — lacking the mathematics for an exact characterization — by a series of “psychological” metaphors, themselves of partly Aristotelian origin: inclination, repugnance, indifference, and violence. […] Galileo’s conception of the Sun’s motion is necessarily hesitant and ambiguous, and he was wary of flatly stating general principals. But one can perceive here the rough outline of what would become Newton’s first law of motion.

What Aristotle, and thus Galileo lacked was the basic material taught to high schoolers and undergraduates in science and mathematics.

It is certainly true that these were exciting and dangerous. Scientific progress has a particular quality of often looking more exciting the further back you stand from it. But the actual doing of science is in the close up; in the detail. And when you do look closer at the danger, at least in the case of Galileo, it looks far more grim than it does exciting.

June 27, 2022

The Final Word

It is the mark of clear thinking and good rhetorical style that when I start a sentence, I finish it in a suitable, grammatical fashion. A complete sentence is synonymous with a complete thought. In the world of AI, completing sentences has become the starting point for the Large Language Models which have started talking back to us. The problem that the neural networks are quite literally being employed to solve is “what word comes next”? Or at least this is how it was explained in Steven Johnson’s excellent Times article on the recent and frankly impressive advances made by the Large Language Model, GPT-3, created by OpenAI.

As is apparently necessary for a Silicon Valley project founded by men whose wealth was accrued through means as prosaic as inherited mineral wealth and building online payment systems, OpenAI sees itself on a grand mission for both the protection and flourishing of mankind. They see beyond the exciting progress currently being made in artificial intelligence, and foresee the arrival of artificial general intelligence. That is to say that they extrapolate from facial recognition, language translation, and text autocomplete, all the way to a science fiction conceit. They believe they are the potential midwives to the birth of an advanced intelligence, one that we will likely struggle to understand, and should fear, as we would a god or alien visitors.

What GPT-3 can actually do is take a fragment of text and give it a continuation. That continuation can take many precise forms: It can answer a question. It can summarize an argument. It can describe the function of code. I can’t offer you any guarantee that it will actually do these things well, but you can sign up on their website and try it for yourself. The effects can certainly be arresting. You might have read Vauhini Vara’s Ghosts in the Believer writing about her sister’s death. She was given early access to the GPT-3 “playground” and used it as a kind of sounding board to write honestly about the death of her sister. You can read the sentences that Vara fed the model and the responses offered, and quickly get a feeling for what GTP-3 can and can’t do.

—

It will be important for what I am about to say that I explain something to you of how machine learning works. I imagine that most of you reading this will not be familiar with the theory behind artificial intelligence, and possibly intimidated. But at it’s core it is doing something quite familiar.

Most of us at school will have been taught some elementary statistical techniques. A typical exercise involves being presented with a sheet of graph paper, with a familiar x and y axis and a smattering of data points. Maybe x is the number of bedrooms, and y is the price of the house. Maybe x is rainfall, and y is the size of the harvest. Maybe x is the amount of sugar in a food product, and maybe y is the average weekly sales. After staring at that cluster of points for a moment, you take your pencil and ruler and set down a “line of best fit”. From the chaos of those disparate, singular points on the page, you identified a pattern, a correlation, a “trend”, and then impose a straight line — a linear structure — on it. With that line of best fit drawn, you could then start making predictions. Given a value of x, what is the corresponding value of y?

This is, essentially, what machine learning and neural networks do.

The initial data points are what is referred to as the training data. In practice however, this is done in many more than two dimensions — many, many more, in fact. As a consequence, eyeballing the line of best fit is impossible. Instead, that line is found through a process called gradient descent. Taking some random choice of line as a starting point, small, incremental changes are made to the line, improving the fit with each iteration, until that line arrives in a place close to the presumed shape of the training data.

I say “lines”, but I mean some kind of higher dimensional curves. In the simplest case they are flat, but in GPT-3 they will be very curvy indeed. Fitting such curvy curves to the data is more involved, and this is where the neural networks come in. But ultimately all they do is provide some means of lining up, bending, and shaping the curves to those data points.

You might be startled that things as profound as language, facial recognition, or creating art, might all be captured in a curve, but please bear in mind that these curves are very high dimensional and are very curvy indeed. (And I’m omitting a lot of detail). It is worth noting that the fact we can do this at all has required three things: 1) Lots of computing power 2) Large, readily available data sets 3) a toolbox of techniques and heuristics and mathematical ideas for setting the coefficients that determine the curves.

—

I’m not sure that the writing world has absorbed the implications of what this all means. Here is what a Large Language Model could very easily do to writing. Suppose I write a paragraph of dog-shit-prose. Half baked thoughts put together in awkwardly written sentences. Imagine that I highlight that paragraph, right click, and somewhere in the menu that drops down there is the option to rewrite the paragraph. Instant revision with no new semantic content added. Clauses are simply rearranged, the flow is adjusted with mathematical precision, certain word choices are reconsidered, and suddenly everything is clear. I would use it like a next level spell checker.

And that is not all: I could revise the sentence into a particular prose style. Provided with a corpus of suitable training data I could have my sentences stripped of adjectives and set out in terse journalistic reportage. Or maybe I opt for the narrative cadence of David Sedaris. So long as there is enough of their writing available, the curve could be suitably adjusted.

In his Times article, Johnson devotes considerable attention to the intention and effort of the founders to ensure that OpenAI is on the side of the angels. They created a charter of values that aspired to holding themselves deeply responsible for the implications of their creation. A charter of values which reads as frankly, and literally hubristically, as they anticipate the arrival of Artificial General Intelligence, while they fine tune a machine which can convincingly churn out shit poetry. Initially founded as a non-profit, they now have birthed the for-profit corporation OpenAI LP, but made the decision to cap the potential profits for their investors: Microsoft looming particularly large.

But there was another kind of investment made in GPT-3. All the collected writings that were scraped up off the internet. The raw material that is exploited by the gradient descent algorithms, training and bending those curves to the desired shape. Ultimately, it is true that they are extracting coefficients from all that text-based content, but it is unmistakable how closely those curves hew, in their abstract way, to the words that breathed life into them. They actually explain that unfiltered internet content is actively unhelpful. They need quality writing. Here is how they curated the content in GPT-2:

Instead, we created a new web scrape which emphasizes document quality. To do this we only scraped web pages which have been curated/filtered by humans. Manually filtering a full web scrape would be exceptionally expensive so as a starting point, we scraped all outbound links from Reddit, a social media platform, which received at least 3 karma. This can be thought of as a heuristic indicator for whether other users found the link interesting, educational, or just funny.

(From Language Models are Unsupervised Multitask Learners)

For all of Johnson’s discussion of OpenAI’s earnest proclamations of ethical standards and efforts to tame the profit motive, there is little discussions of how the principal investment that makes the entire scheme work is the huge body of writing available online that is used to train the model and fit the curve. I’m not talking about plagiarism; I’m talking about extracting coefficients from text, that can then be exploited for fun and profit. Suppose my fantasy of a word processor tool that can fix prose styling becomes true. Suppose some hack writer takes their first draft and uses a Neil Gaiman trained model to produce an effective Gaiman pastiche that they can sell to a publisher. Should Gaiman be calling his lawyer? Should he feel injured? Should he be selling the model himself? How much of the essence of his writing is captured in the coefficients of the curve that was drawn from his words?

Would aspiring writers be foolish not to use such a tool? With many writers aiming their novels at the audiences of existing writers and bestsellers, why would they want to gamble with their own early, barely developed stylings? Will an editor just run what they write through the Gaiman/King/Munroe/DFW/Gram/Lewis re-drafter? Are editors doing a less precise version of this anyway? Does it matter that nothing is changing semantically? Long sentences are shortened. Semi-colons are removed. Esoteric words are replaced with safer choices.

The standard advice to aspiring writers is that they should read. They should read a lot. Classics, contemporary, fiction, non-fiction, good, bad, genre, literary. If we believe that these large language models at all reflect what goes on in our own minds, then you can think of this process as being analogous to training the model. Read a passage, underline the phrases you think are good, and leave disapproving marks by the phrases that are bad. You are bending and shaping your own curve to you own reward function. With statistical models there is always the danger of “over-fitting the data”, and in writing you can be derivative, an imitator, and guilty of pastiche. At the more extreme end when a red capped, red faced, member of “the base” unthinkingly repeats Fox News talking points, what do we have but an individual whose internal curve has been over-fitted?

It is often bemoaned that we live in an age of accumulated culture, nostalgia, retro inclination. Our blockbusters feature superheros created in a previous century. There is something painfully static and conservative about it all. But what if artificial intelligence leads us down the road to writing out variations of the same old sentences over and over again?

—

In Barthes’ essay The Death of the Author he asserts

We know now that a text is not a line of words releasing a single ‘theological’ meaning (the ‘message’ of the Author God) but a multi-dimensional space in which a variety of writings, none of them original, blend and clash. The text is a tissue of quotations drawn from the innumerable centers of culture.

Given that Barthes was quite likely half-bullshitting when he threw out phrases like “multi-dimensional space”, what I have quoted above is a disturbingly accurate description of the workings of a large language model. But it isn’t describing the large language model. It is a description of how we write. Barthes continues:

the writer can only imitate a gesture that is always anterior, never original. His only power is to mix writings, to counter the ones with the others, in such a way as never to rest on any one of them. Did he wish to express himself, he ought at least to know that the inner ‘thing’ he thinks to translate’ is itself only a ready-formed dictionary, its words only explainable through other works, and so on indefinitely.

Maybe Gaiman would have no more cause to call his lawyer than all the great many writers he absorbed, reading and then imitating in his youth and early adulthood. Maybe if Barthes is to be believed here, the author is dead and the algorithm is alive. Our creativity is well approximated by a very curvy curve in a high dimensional space.

One potential outcome that does not seem to have been considered by our Silicon Valley aristocracy is that the Artificial General Intelligence they bring into this world will be an utterly prosaic thinker with little of actual interest to say. It shouldn’t stop it from getting a podcast, though.

Or maybe Barthes was wrong and maybe Large Language Model will continue to be deficient in some very real capacity. Maybe writers do have something to say, and their process of writing is their way of discovering it. Maybe we don’t have to consider a future where venture capitalists have a server farm churning out viable best-sellers in the same fashion they render CGI explosions in the latest Marvel movie. Maybe we should get back to finishing the next sentence of our novel because the algorithm won’t actually be able to finish it for us. Maybe.

May 12, 2022