An interview about THE FUTURE with Hugo Winning author CHARLES STROSS!
Fumblingly carried out by John Shirley
Charles Stross is masterfully eclectic. He writes science-fiction (he won Two Hugo Awards, a Prometheus Award and a Locus Award, was nominated for many others), alternate history novels (The Merchant Princes series), fantasy (he won a Locus fantasy award for his novel The Apocalypse Codex), and he combines many of these genres, along with social satire and horror and espionage, in his famous Laundry Files series, starting with The Atrocity Archives. My theory is that he actually has several brains in cold storage, which he switches out, somehow, depending on the project.
Stross has also written nonfiction, professionally, about computer science, and he knows what he’s talking about. Currently he’s completing his delightful merging of historical fiction, Lovecraftian horror and somewhat surreal satire, the Tales of the New Management, which commenced with Dead Lies Dreaming and Quantum of Nightmares. It was his futurist thinking brain we wished to consult at Instant Future:
Q. Long ago and far away you got a degree in Computer Science. In your novel Accelerando —an interweaving of discrete stories set in the same universe—the technological singularity is portrayed as essentially inevitable. Your sf is somewhat socially satirical, as well as A-1 primo quality science-fiction, but does the more practical computer scientist lurking inside you foresee a world-changing singularity happening in the 21st century or soon after?
A: the Singularity, as commonly used in SF and transhumanist discourse, is almost certainly a bust. I’ve come to dislike the term intensely over the past decade, as it uncritically subsumes patterns of belief that originated in Christianity: mind/body dualism, the perfectability of intellect, the inevitability of a sudden apocalyptic end to mundane human life brought about by an apocalyptic intrusion of god-like mind — these are all aspects of Christian faith rather than anything with a scientifically testable basis. And they get recycled endlessly in discourse about the Singularity, even before we get to the mind uploading moonshine. It harks back to the writings of the Russian Orthodox theologian Nikolai Fyodorovich Fyodorov in the late 19th century, by way of Russian revolutionary Cosmism.
Fyodorov asked, from a devoutly Christian point of view, what the purpose of humanity was on Earth. He decided that it was only logical that God put Man here in order to bring about some divine purpose, so he started trying to work out a forward-looking teleology. First, he decided, humanity must come together in peace. And what better reason than to pursue a cause everybody can agree on — a struggle to overcome the ultimate enemy, death. We must become immortal: we must learn how to sail between the stars and expand to colonize the entire cosmos (*obviously* that’s what it’s there for!) and then, having become immortal, we must work out how (through science!) to resurrect everyone who has lived and died before us. Finally, by perfecting human nature, humanity gets to build heaven and God shows up. He called this program the Common Cause of Humanity, and it is a strikingly prescient precursor of the ideology of today’s Longtermists … because that’s where they got it.
Fyodorov worked as a teacher and librarian, and among his students was Konstantin Tsiolkovsky, the mathematician who gave us the rocket equation, theorized about the use of cryogenic liquids as rocket fuels, and in turn inspired Robert Goddard and Wernher Von Braun. The space settlers movement also taps into Fyodorov’s Christian transhumanism via Tsiokovsky’s writing.
Now, there’s a deep vein of this theological kryptonite embedded in the Enlightenment era concept of “progress” — for the very concept of progress is a peculiarly Protestant heresy — and western civilization and the economic theories it runs on are largely built atop it. The Enlightenment also brought us a package of other beliefs, notably a touching faith in the explanatory power of rational thought. We all come with cultural baggage — we’re marinating in a sea of beliefs which we simply don’t question, much as fish don’t question the presence of water — and although it has become increasingly respectable to question supernatural or religious belief systems, it looks from where I’m standing as if many self-avowed atheists and rationalists are actually replacing the religion they rejected with an elaborate framework of beliefs that are structurally indistinguishable from it.
I’m not Christian. But I was raised Reform Jewish, which meant I was unable to avoid being confronted by Christian beliefs — many of which look batshit crazy, to my outsider’s eyes — on all sides. And when I looked at the Extropians and Longtermists and Singularitarians what I saw was simply refactored Christianity, minus God and Jesus, with a whole bunch of new saints bolted on top. (Ray Kurzweil! Satoshi Nakamoto! Konstantin Tsiolkovsky!) And this similarity inclines me to adopt the same wary skepticism.
I cannot refute the possibility that a hard take-off AI singularity will take place in the next few decades, any more than I can refute the possibility that the Rapture will take place at 9:47am next Tuesday and all the Elect of Jesus will go a-flying up to Heaven. But I think either event is about equally likely.
Q. Will the pace of change overwhelm us? I seem to perceive, behind many of your novels, a writer conflicted about technological advancement; not against it, certainly no luddite, but concerned about its nature. It would seem that we need that advancement—but we’ve failed to develop a protocol for advancing technology intelligently. For one thing, a technology that pollutes is only half-invented. This seems clear in the age of anthropogenic climate change. Should we slow the pace? Can we?
A: I think, going by the news headlines, the pace of change has *already* overwhelmed us. The Tofflers made this case fairly well in their book *Future Shock* back in the 1970s, and that was in a then-stable media environment that wasn’t polluted with memes generated by bad actors (eg. state level disinformation agencies) and chatbots (often just trying to sell something — Ivermectin as a cure for COVID19, for example).
One problem is that we’re nearing the crest of a sigmoid curve of accelerating advances in a new technological area — computing, networking, and information processing. It seems unlikely progress on miniaturization of semiconductors will proceed for many more generations (our densest integrated semiconductor circuits already have tracks and other features on the order of a hundred atoms wide: it’s hard to see how we can shrink mechanisms below the atomic scale). So, just as progress with steam locomotion had tapered off by the 1920s after a brisk acceleration from roughly 1790 through 1870, and aviation surged from the original Wright Flyer and its contemporaries around 1900 to the SR-71 and Boeing 747 by the early 1970s but subsequently stopped getting bigger or faster, we’re approaching an era of consolidation and very slow incremental gains in our IT. People are now exploring the possible ways of monetizing the technologies we’ve acquired over the past few decades, rather than making qualitative breakthroughs. I first saw a virtual reality headset and interface in use at a conference in the early 1990s; the fact that Apple are apparently bringing one to market this summer, and Meta (aka Facebook) sank billions — evidently fruitlessly — into trying to commercialize VR over the past few years, should be a huge warning flag that some technologies just don’t seem to be as useful as people expected.
We’re also suffering from information pollution in new and unexpected ways. Colour cycling light bulbs seem like a neat toy, and when we’ve got cheap LEDs they’re easy enough to build. And it turns out that mass-produced microcontrollers (embedded microprocessors with a tiny amount on onboard storage and i/o channels) are cheaper to bodge into a $10 light bulb than designing custom circuitry. So we now have millions of light bulbs with wifi and embedded microcontrollers running obsolete, unsupported versions of Linux with unpatched security exploits, which can be hacked and co-opted to join botnets or waste electricity mining bitcoins or cracking passwords so they can act as staging posts for ransomware targeting your laptop. Worse: some joker decided to add speakers to the light bulbs so they can play music. Which sounds fine until you think in terms of malware that looks for people known to suffer from mental illness and not filling their prescriptions so that they can be targeted with suggestions generated by chatbot to motivate them to, well, who would you like your army of unmedicated malware victims to target today?
And that’s without getting into the utility of modern social networks for snake-oil hucksters and time-share salemen and televangelists and neo-nazi recruiters.
We’ve created a hideous grifters paradise, where everybody needs to know stuff that only network security administrators needed to be aware of a couple of decades ago, and made it a terrible time to be a paranoid schizophrenic.
Q. Is there technology arising in our time that took you by surprise? I predicted self driving cars but was surprised they came so soon. I was somewhat taken aback at the arrival of 3D printing. It seems possible that 3D printing may strikingly transform our economy, for better or worse. What do you think?
A: I was blindsided by mRNA vaccines. That stuff is *amazingly* cool, and it came out of the early-stage R&D lab to billions of vaccine doses in arms in a matter of months thanks to the COVID19 pandemic (which is still with us). It *shouldn’t* have been a surprise, but of course I was looking in the wrong direction. It turns out that anything to do with genomics or proteomics is incredibly compute-intensive, so since about 1990 we’ve been seeing a positive feedback loop between Moore’s Law (computing cycles getting exponentially cheaper) and our understanding of genetics (which is vastly more complex than we realized, even in the 1980s). The original Human Genome Project was scoped to take a decade and cost $200M to produce a comprehensive sequence for one individual. As it happens, the last 90% of the work all took place in the last two years and it finished early and significantly under-budget … then the cost kept falling: if the UK’s national health service wasn’t under attack by neoliberal disaster capitalists, I would expect it to begin offering routine personal genome sequencing to the entire population by 2030, simply because of the cost savings they could make if people at high risk of hereditable conditions could be identified and offered preventative treatment before the cancer or dementia or cardiovascular disease got its claws in.
3D printing of metals is also something I didn’t expect, or thought would be horrendously expensive — not so cheap that Rocket Labs would be printing entire space launchers!
And the biggest disruption of all is photovoltaic cells. They (like batteries) are on an exponentially declining price curve this decade. Until the 1990s, it generally cost more in energy to manufacture a solar cell than it could generate during its working life. But now the things are so cheap that installing new solar farms is cheaper than burning the same wholesale priced amount of coal in *already amortized* coal-fired power plants. We print the things on plastic film that weighs about as much as printer paper, and we make it by the square kilometre. We may well end up in that 1970s scifi future powered by orbital powersats, only with thin-film panels delivered be fully reusable rockets running on synthetic fuels (I’m looking at SpaceX’s Starship here, once they get past the initial problematic development phase — which is by no means guaranteed).
Q. Are AIs as dangerous as they seem? Clearly, they could be used for social manipulation in dangerous ways, but I for one doubt they’ll ever become truly sentient and self-aware and self-deterministic. Am I wrong?
A: AIs do not exist at this time. The things being sold as AI are not intelligent, they’re just statistical models based on gigantic aggregated data sets. They produce amazing results, but there’s no theory of mind there. And *that* is the true danger of AI, at least today!
In the 1980s one sure sign of a grifter was them pitching a technology at you with the prefix “cyber-” bolted on board. (These days the only folks who still use “cyber-” seem to be the military. And, uh, us writers.) In the 2010s the grifters were all aboard the blockchain and cryptocurrency, aka “crypto” (much to the intense irritation of actual cryptographers, who are basically applied mathematicians and security experts). When the cryptocurrency bubble finally burst in 2021/22, the grifters promptly jumped aboard the next money-making bandwagon, the aforementioned statistical models, which use a drastically simplified model of how neural networks work — so *of course* they had to put lipstick on the pig and call it “artificial intelligence”.
Which will kill people in large numbers, sooner rather than later, as a result of real humans taking the buzz-word at face value. Just like Tesla, who labeled their advanced cruise control “autopilot”, and thereby bamboozled the driving public (who don’t realize that on an airplane, an autopilot is just three-axis cruise control and must be supervised by a human pilot at all times) into taking their hands off the wheel and their eyes off the road.
Q. The better aspects of civilization seem to be at grave risk in our times. Our abuse of the environment, our runaway, basically-19th-century/ early 20th energy sourcing, like a senile old man with an AR15, is out of control. This seems driven by a barely regulated, greed-hysterical global economy. Charles, if you had a magic wand that could give you the power to reconfigure the world political and economic order, and if you were feeling megalomaniacal enough to do it, how would you reconfigure it? What changes would you make?
A: WHEN I EMERGE AS SUPREME PLANETARY OVERLORD I WILL— (maniacal laughter)
Honestly, I don’t know. It feels like a big chunk of the problem is that the oligarchs are panicking as we near the end of the petrochemical fuel age and confront the anthropogenic climate change hang-over, so they’re trying to build survivalist bunkers *or* pump all the remaining oil *or* kill their perceived enemies (in the age of social media and automation, that means the former working/producer classes).
It feels as if we’re being confronted by an emergent Fascist International who are coordinating and sharing tactics, but I’m not sure it didn’t exist already in earlier decades — remember the Italian masonic lodge P2? Or the John Birch society? Or (in the UK) the Economic League?
In my more depressive moments I cheer myself up by fantasizing about Earth being invaded and forcibly occupied by Iain Banks’ Culture, his spacefaring fully automated luxury gay space communist utopia. But then I have to remind myself they’re just SF.