In recent months, I have been exploring books published by academic presses, and while poking around the website of McGill Queen University Press I discovered the book From Literature to Biterature Lem, Turing, Darwin, and Explorations in Computer Literature, Philosophy of Mind, and Cultural Evolution by Peter Swirski. I was intrigued and wrote to Mr. Swirski to ask for an interview. He very courteously consented and this is the result.
You say of your book, “…my explorations, the first of their kind…” Hmmm, not exactly shy about the value of your book are you? Could you tell us what you mean by that?
Unlike most cultural future gazers, whose gaze is actually fixed on the past or the present, I’m trying to explore the ultimate future of literature by exploring the nature of beings who will create it. If my scenario is correct, “biterature,” as written by computer authors or “computhors,” will be a manifestation of the beginning of the end of the cultural world as we know it. In that sense, From Literature to Biterature is really a book about our human future in the age of thinking machines.
One of the first things that I found startling in your book is the statement that, “…by dint of typing it, I have done my bit to make the word ‘computer’ come up in written English more frequently than 99 per cent of all the nouns in the language.” How did you come up with that figure and what other nouns have comparable rates of frequency in English? Is English more computer-reference heavy than, for instance, French, German, or Korean?
Cluttered with junk as it is, the Internet has some useful resources such as WolframAlpha search engine (one of the crunchers behind Microsoft’s Bing). That’s where I got this particular statistic from. Consider, however, the following sentences:
My Mac is down.
This is a great system.
The processor works like a dream.
It’s a pretty good machine.
Look at my new baby, cost only 900.
None uses the word “computer,” yet all talk about computers. So, like with every statistic, mine should be taken with caution — in this case, revised upward (this ability to refer nonreferentially holds for every language on earth). I’ve no idea about French, German, and Korean languages, but I’d be extremely surprised if their stats were categorically different.
One of the main premises of your book is that in the foreseeable future, computers will become capable of creating works of literature. Let’s talk about what you mean by “literature.” Do you mean literature as in the works of Henry James, Charles Dickens, George Eliot or even J.K. Rowling or Dan Brown? Is that really a problem? Won’t it make life easier for publishers (no royalties to pay, no literary agents to bother with)? And if readers enjoy the books produced by computers, isn’t that all to the good?
This is actually several questions and each the size of Mt Everest. So in the briefest of briefs:
- From Literature to Biterature employs the term “literature” broadly to encompass fiction, biography, philosophy, and other forms of writing (nonfiction is no less a form of creative writing than fiction is). As I argued in another book, From Lowbrow to Nobrow, popular and canonical literatures are all species of literature (much of the canon IS or WAS popular literature when it was created). Dan Brown is a spectacularly ungifted robot.
- No problem at all.
- Computhors, i.e. thinking machines, will be independent entities with their own desires and their own agenda: it’s a mistake to think of them as causally dependent on us in any way. Instead, they will behave in the same way we do: going after their goals in the manner that benefits them.
- Absolutely: we may of course issue warning labels (“this book may be harmful to you as it was created by a machine”) but, like with warning labels about explicit lyrics in rap, this may only titillate buyers.
Do you think some genres are better suited to production by computhors than others? Sports reporting, maybe?
Not at all. Your question is really about today’s parsing systems which, within their limitations, do a decent enough job to be employed by, for example, all major news agencies. In contrast, computhors will be able to create any kind of literature — or anything else, for that matter — and do so spontaneously, without human causal input.
You say that your book is an adventure story of the mind. When you were a kid, did you like adventure stories? What would you consider a great adventure story? Captains Courageous? Treasure Island? How is your book like an adventure story?
Who doesn’t love a good adventure! As a child, I used to devour Jules Verne, Stanislaw Lem, Beowulf and the Norse Sagas, The Sea-Wolf and Jack London’s South seas stories, and thousands of others that fired up my imagination. Some of them I outgrew; others, like Lem and Beowulf, are still my companions.
I prefaced From Literature to Biterature with a lovely quote from Marvin Minsky in which he disclaims footnotes and the entire academic apparatus for the sake of freeing the imagination. It is in this very sense that my book is an adventure story of the mind (I tried to write a book without footnotes but the press would not let me, so there are a few left).
You refer to Katherine Hayles’s 2008 book, Electronic Literature, as a “cheerleading overview of computer-assisted writing” and “a survey of second-order mashing protocols of yesterday.” Would you like to elaborate on that?
The book is clearly enthused by the meager, in my opinion, literary payoffs of writers employing computer protocols to mash their texts. More to the point, this has been going on for decades — I give an overview of the scene in Chapter One. In contrast, From Literature to Biterature goes after the ultimate future of books and authors — one that looks nothing like the past or the present.
You refer to Asimov’s “Three Laws of Robotics” as “asinine.” Why? Are such laws in general silly or are Asimov’s, in your view, unusually stupid?
They are literally nonsensical since they are self-contradictory. Asimov knew it very well but cleverly wrung a lot of narrative mileage out of playing (not always fairly) with the literary conventions he established. Alas, his exegetes often confuse playful storytelling with science fact.
You write that in the place of bad science fiction we need good science and good fiction. Do you regard most science fiction as bad or just some of it? Do you read much science fiction? Whom would you consider a purveyor or bad science fiction? What do you think of Philip K. Dick, Ursula K. Le Guin, Ray Bradbury, and Margaret Atwood?
I used to teach science fiction for years and together with my students had plenty of fun poking holes in every–and I mean every–definition of the genre in existence. In general, academic taxonomy strikes me as a species of taxidermy. But, since you asked, let me put myself in the line of fire:
- Dick was a science-fiction visionary but an extremely sloppy one.
- Some of Le Guin’s ideas are as inspired as her prose isn’t.
- Bradbury was not writing science fiction as he himself stated on numerous occasion (I cite one such disclaimer in my book Of Literature and Knowledge).
- Atwood is an excellent writer but not of science-fiction.
Could you please give an example, if one exists, of what you would consider a “third-order computhor?” For example, you have never met me. How do you know for sure I am not one? You make it clear in your book that your readers would be hard pressed to prove that Peter Swirski is not an android. Was that fun to write?
There are no emphatically computhors around yet, although they will come to pass, of this I am sure.
As for your other question, I do NOT know for sure that you are not an android, just like you do NOT know that I’m not. Matter of fact, I AM an adroit android — I confess to it on the pages of From Literature to Biterature. Don’t believe me? Prove me wrong!
In general, it is our evolutionary inbuilt module called Theory of Mind (ToM), plus common sense, that help us override such solipsism/skepticism.
I am looking at a digital copy of your book (which will also be available as an e-book) and it says, “Right now, the patterns inscribed on this page cause you to flex your muscles so as to hold a bound sheaf of pages in your hands…” A computer will not have the lingering attachment you seem to have to printed books. Will its inability to experience sensations as humans do affect the computhor’s prose style and powers of characterization and scene setting and atmosphere creation (as opposed to plot creation)? Computers can’t smell can they, as Orwell could? Or can they be programmed to act as if they can?
Computhors will NOT be programmed by us. They will be programming themselves — if the idea of programming makes any sense at that stage of their development/evolution. We’re not talking about machines in our sense of the word, but about nonhuman sentient beings. Do we program other people that we come into contact with: children, friends, employees? If you answer “yes” — and most people would balk at that — this is the only sense in which we would be programming computhors.
If they choose to experience the world as if they were devoid of optical, infrared, acoustic, magnetic, or any other feedback, and write about it, it will be their prerogative. Please note: “independent” does NOT automatically mean “hostile” or even “indifferent”–we are not talking third-grade science fiction here.
Computers also don’t age in the way humans do or become disabled in the way humans do. Will that affect how they write? You make the point that they can be programmed to seem to age at any rate the programmer chooses and that they can do so much more than we can (survive at great depths beneath the sea, operate in darkness in ways most humans cannot, etc.) in some respects. But who will want to read literature so disconnected from the average human experience? After all, even now science fiction appeals to a substantial but still niche audience.
Precisely: computhors will most likely not write for us, except for atavistic, egotistic, or prestige-based reasons. If they create art, they will create it for themselves and quite possibly do it in a way of which we will not even be aware — unless they decide to let us know.
So could there be a Conan Doyle computhor type and a Virginia Woolf type computhor, or any other author we can imagine?
Absolutely. Computhors will have personalities, styles, creative differences, and so on. Then again, one computhor could easily harbour a Conan Doyle and Woolf personalities and/or styles within and switch or combine them at will. In reality, however, computhors will almost certainly have different types of personalities from us.
You express concern about the rise of super-fast trading in financial markets and the fact that the pace of transaction is dangerously fast. Trading was halted on NASDAQ for several hours on August 22, 2013 due to what appears now to have been a flood of data that seems to have gummed up the whole system. I am not quite clear what problems like that have to do with biterature. Could you explain?
Causal independence and creative autonomy will emphatically not arise in isolation, just like you cannot have a human being capable of creating magnificent fiction and not lying in real life. Intelligence is a wave spreading in all directions at once. If you have systems capable of spontaneously creating literature, you have systems capable of deliberately crashing a market — or, for that matter, propping one that is on the verge of a crash (of which we would likely not even be aware).
You write that computhors would be “black boxes” compared to the vast amounts we know about the lives of some authors. But isn’t that all to the good? Wouldn’t it be sort of nice to be able to read poetry by a computhor that had no biographical baggage to speak of?
[Smiling] Authors try to say only as much as it suits them. It will be no different with computhors. They will be as revealing or reticent as suits them. As for biographical baggage, as readers or scholars we can always choose to ignore it. However, the computhor’s provenance (or life history) may be essential to understanding its particular artistic vision, for example, if it chooses to create as a “blind” artist.
Could computhors be used for nefarious purposes, like acting like trolls on blogs or in Twitter or touting books or other products in online forums? Has this already happened?
The question is misleading: computhors could not be used in the sense that we use industrial cranes to lift heavy cargo. Cranes have no volition and personality: thinking machines will. So, to rephrase: have computers been used for mass-trolling or blogging? I’ve no idea, although judging by the quality of many blogs and posts, they may have Will computhors troll and blog? Perhaps, since they will certainly be capable of it. But remember that these will be entities capable of evolving millions of times faster than us, entities that never get tired, entities that can in effect be everywhere on the planet at once, entities that will have goals and drives that will be forever unknown (and most likely unknowable) to us.
You joke about a mythical periodical called, “The New York Times Biterary Review.” But is that really so far-fetched and have you considered starting such a periodical? Would you invest money in one? After all, if we are entering a world in which computhors churn out the vast numbers of works you predict, won’t we need periodicals that review them? And many of those reviewers could be human beings and could get paid real money (and could be truthful in their reviews because you can’t libel a computhors — or can you?).
I’m currently writing a sequel to From Lowbrow to Nobrow entitled From Highbrow to Nobrow, which deals head-on with the multitude of questions you ask. Most of them have a common denominator, and it is numerical. The number “zettabyte,” which is slowly beginning to approximate the totality of cultural information created by our species, is so enormous that it is in many ways incomprehensible. This quantitative expansion is completely upturning the cultural rules that have been in place for millennia. As cultural producers and consumers we have created a runaway universe of information.
You write quite passionately about the power of literature written by human beings (as opposed to computhors) to affect us emotionally. What novels have you read in the last year? What literary works would you recommend every person try to read before she dies?
I’m a voracious reader — and, as I get older, re-reader. This year I’ve read or re-read Lem, Beaton, McMurtry, Heller, Collins, LaHaye and Jenkins, McBain, Semonov, Van Vechten, Shakespeare, Connelly, La Plante, Patterson, Alten, Chandler, Cain, Alex, Savchenko, Poe, Pope, Cooper, Sistah Souljah, Anonymous, Bateman, Buckley, Percy, and dozens of others I don’t remember.
I would not presume to foist my literary idiosyncrasies on other readers. But, from my personal collection, I’d probably take these ten fictions to a desert island:
Shakespeare, Collected Works
Lem, His Master’s Voice
Lem, A Perfect Vacuum,
McMurtry, Lonesome Dove
Capek, War with the Newts
Ellroy, L.A. Confidential
Dos Passos, USA
Perez-Reverte, The Club Dumas
McBain, 87th Precinct cycle
You make the fascinating point that if something is unintelligible it takes time and resources to unravel its meaning. Could you relate that to the current controversy over the NSA’s slurping up amounts vast amounts of information from average citizens and its arguments that this is “only” metadata and that its acquisition of such is imperative for national security? Do you wish that you had been able to discuss Edward Snowden in your book?
One way of saying that something is unintelligible (as opposed to being only supremely difficult to decipher and/or understand), is that it would take an infinite amount of time and resources to render it intelligible. Biterature may be such a phenomenon: created by our creations, it may be for all purposes completely unintelligible to us (hence we would never actually know that it is a form of literature).
I’ve written several books on American politics and politricks, among them Ars Americana, Ars Politica, then American Utopia and Social Engineering, and forthcoming War on Errorism. To readers familiar with them, the current hoopla over illegal spying on the nation living allegedly in the land of the free is not a surprise.
As for Snowden, he should get a Congressional Medal of Honor. On the other hand, as an intelligence analyst, he was incredibly naive to think he could blow the whistle on Uncle Sam and walk away unmolested.
You write, “A thinking computer will be just as much of a person — albeit a non-person — as you or I.” Does that mean I can’t throw it away?
Everything depends on your meaning of “can’t”. Can you squash a worm? Yes. Yet most people tend to step over worms. Can you kill a human being? We all have that capacity, yet tend not to exercise it and find even contemplating the act disgusting and disturbing. A thinking computer will be a nonhuman person that, yes, you would be able to throw it out of the window. But would you? Odd as it sounds, people sometimes take the trouble to look for a “good home” when they have to give away their favorite sofa, a rather far stretch from a thinking being.
Whom do you see as ideal readers of your book? Philosophers? Those who study literature and write about it? The educated general reader who wants some idea of what is happening vis-à-vis the impact artificial intelligence is having on cultural life?
Hopefully all of the above. I try to write what I call beach books for intellectuals — highbrow content, lowbrow style — and From Literature to Biterature is no different. It’s a book for everyone with a hungry mind who cares about the future of books and the future of humankind (and likes to be entertained while being enlightened). I probably just described every reader of Critical Margins.
You write, “…the environment will have become more intelligent than its inhabitants.” Could you define what you mean by “environment” in this case?
Even our relatively primitive civilization would collapse instantly if all computation stopped for some reason. The future involving thinking machines will only make us more dependent on this computational environment. I weave several playful scenarios in the last part of From Literature to Biterature to show how the decision to surrender ourselves in this environment will have been taken out of our hands.
Do you think the recent purchase of The Washington Post by Jeff Bezos suggests anything of significance about the way news will be produced in coming years?
For Amazon — an overpriced, overrated, and reader-unfriendly corporation — it could be merely a shrewd political move to secure an unparalleled lobbying outlet in the heart of Washington.
Is the ability to play chess well really a good measure of human intelligence?
No. Deep Blue and X3D Fritz, two programs that did so well against arguably the best human chess player ever, are morons (strictly speaking not even that) away from the chessboard. It is like saying that coin sorters are more intelligent than us because they can outdo humans in sorting coins.
You started your professional life in literary studies, but this book is greatly concerned with philosophy and matters of logic and neuroscience. What has been the reaction to your book so far among scholars in all of those fields?
Literature and the study of it are my life and my passion, not least because they take me to so many places outside literature. As for From Literature to Biterature, outside literary venues, it got me invited to a panel on BBC World Service with a roboticist and cognitive scientist; it received a glowing endorsement from perhaps the greatest living logician Jaakko Hintikka; and it will be reviewed in a couple of top journals of philosophy — not bad for a book that’s not even out yet. Having said that, I would love to hear from all and any readers of From Literature to Biterature at firstname.lastname@example.org.
Thank you for your time.
My pleasure, Hope, and thank you for this perfect mix of profound and playful questions.