While educated readers typically consider science fiction as more of a lens on our present world rather than literal prediction or futurology, it’s interesting to look closer and notice how badly inaccurate science fiction’s imagined futures have been.

It’s not that we need to hold science fiction to a high standard of accuracy about the future or “improve” it, but it’s remarkable how consistently wrong most of it is in exactly the same direction. That matters, as science fiction has emerged as a large influence on the most powerful businesses and technologists today. They may have learned all the wrong lessons. [1]

I’ll just give some examples that come to mind where even with reams of science fiction produced on the topics, we’ve found ourselves surprised when the future arrives. Exceptions are notable because they are so rare, and right in the same ways.

Computers

Early science fiction stories about computers mostly imagined their future versions as monolithic, expensive and more advanced versions of 1940s computers. [2] You see this all the way into the 1970s. The few that imagined a computer in every home are remembered precisely because they were so unusual. [3] When all at once in the 1980s you could buy your own personal computer at Sears and Radio Shack for a few weeks’ pay, it wasn’t a future anyone had imagined, not really. Asimov imagined computers becoming incredibly effective, but not smaller and cheaper. His thesis was along the lines of “What if the Univac gets really, really good?” and logically worked out some entertaining and frightening situations.

Microcomputers exploded onto the scene in the late 70s and early 80s. Fiction adjusted rapidly with Cyberpunk. It’s like the 1980s but doubles down on the neon and hackers. Cyberpunk mostly failed to anticipate further miniaturization to anything but computers, and therefore missed the pocket cellphone, and all the plot devices it would eliminate.

Gibson almost got there: He was inspired by an Apple IIc ad (the IIc had a handle so you could carry it around like a briefcase). [4] He correctly understood the IIc was just the start: Go right to small “decks” you can jack straight into, no screen required. People could meet up in cyberspace. Yet, there were also payphones in Neuromancer and Count Zero. He shows us plenty of high-tech miniature spy gear and medical implants like artificial eyes – you see this stuff in most cyberpunk and generally in science fiction starting in the 80s. It’s all exotic and expensive and secret. Stuff that’s scarce is interesting in a story. Miniaturization was noticed, but mass production, overlooked.

Social Media and Advertising

Social media can only exist in a world with universal internet access and individual private access. It also needs financial support. The idea that the money comes from ads may have been just a little too cynical for most writers.

Again Gibson came close in spirit. We don’t have simstim (sensory stimulation that skips your eyes and ears and touch and goes right to the brain), but how it’s used sure feels realistic. Count Zero and Mona Lisa Overdrive have simstim actors who precisely resemble YouTube and Instagram influencers down to their quasi-journalist roles, sometime product promoters having parasocial relationships with their fans. Kids want to be simstim stars when they grow up. [5]

What’s missing is the overwhelming flood of advertising aimed at users and the sheer amount of time spent on social media. The nature of Gibson’s simstim prevents total absorption except by addicts. Normal people have to go to work, eat dinner, etc. You can’t do both at once. Phones, on the other hand, being less immersive, allow us to walk around paying only half attention to everything; it’s not all or nothing. That’s probably a reason VR hasn’t really taken off. Until Meta VR glasses are true simstim from Neuromancer you’re not going to convince most people to spend a lot of time in them.

Philip K. Dick almost uniquely imagined the hell we’d create for ourselves through the combination of miniaturization plus advertising. In The Simulacra, [6] advertising reaches us in the form of “Commercial Flies,” autonomous, mass-produced, tiny drones that crawl through cracks in windows and doors, sticking to surfaces and broadcasting ads loudly. People treat them like nuisances and swat them routinely. The theme of harassment by advertisers comes up routinely in his fiction.

I point these out because they are well remembered, but quite rare for their time, the exception, not the rule. Citing all the stories by their contemporaries that utterly failed to grasp the basic principle of advertising and mass production would be too exhausting, but here are some examples. [7]

Artificial Intelligence

AI by way of LLM coding agents has taken software developers by surprise – we’re the first ones the machine can replace, not the last ones as we always believed would be the case. Education was completely unprepared for the rapid spread of AI technology, and rather than personalized tutors we have paper mills in students’s pockets. It’s like nobody imagined what would happen if machines developed some kind of intelligence. Of course that’s not so; there are piles of science fiction out there. What did they miss?

The big difference between AI in most of those books and reality is just how cheap and ubiquitous AI is, and how quickly it happened. Universal access to the internet to deliver the AI “product” makes this possible, and few writers could really envision an internet, let alone what it would unleash.

Most stories envision AI as a single entity, sometimes confined to a singular physical machine. Gibson may have gotten closest along with Vernor Vinge in understanding AI as pure software, but they didn’t see the front and center total commercialization side – they imagined AI emerging stealthily somewhere on the internet, something haunting backwater corporate mainframes or elite research supercomputers without their operators’ knowledge. [8] Again, like with computers, in stories the new thing is something scarce and secret, not mass-produced.

Vinge and Gibson could apply logic and work from first principles to convince us of bizarre for the time scenarios, but these were set within a larger universe that wasn’t so different from their present.

In Neal Stephenson’s “The Diamond Age,” the upper crust gives their children “primers,” AI-powered books that intelligently interact with the kids to teach them. These are expensive and rare. The truly expensive ones are guided by a real person on the other end (the AI is said to be inferior to a real human tutor). The “Young Lady’s Illustrated Primer” makes for a compelling goal if you imagine you might actually be capable of building it.

We didn’t spend time worrying about how kids would still learn anything once AI arrived, because nobody supposed it was plausible that every kid would have basically unlimited access to their own AI, and have it in their pocket to help them cheat on essays. And, the AI is simultaneously capable of complex intellectual work but has no will of its own to refuse or look out for the child’s best interest.

The unexamined assumption was that AI would be costly, possibly secret, and certainly fairly exclusive. Costly it is, but the nature of the internet and computers makes it infinitely copyable and available. Plus, giant corporations are trying their best to make everyone use AI whether they want to or not.

Immortality

Life extension or immortality is another case. Science fiction normally assumes some advances, maybe enormous strides, will take place. But these will be available only to the very richest. Sometimes there’s some serious limitation, like you have to harvest organs or undergo some very expensive personalized procedures.

Both serious hard science fiction and less rigorous space opera stories typically depict medical science advancing incrementally, offering longer life to only the most privileged. The examples are too numerous to list. [9]

We haven’t quite hit a massive breakthrough, but if we do there’s no reason to think science fiction has a lot to tell us about what comes next. As a small-scale example we have GLP-1 drugs and related treatments that can not just take off weight like nothing else (well I guess there’s always fasting) but will improve health on many dimensions. They’re just getting started and rapidly becoming available to huge numbers of people at once. In most near-future science fiction this is the sort of drug that only a select few would have access to.

Since long healthy life is something we’d value highly, it makes intuitive sense when life extension technology is depicted as valuable and hard to get.

What if that’s not so, and the solution to living to 200 is as simple as a GLP-1? The story flips from the expected have vs. have-not framing to something nobody expected. AI may help us out here. I sure don’t know what advances may come soon in life extension, but I wouldn’t be surprised to find out it’s something relatively cheap and easy. The story changes from life extension as a valuable and coveted thing, to a story asking “what do we do now that everyone has it?” [10] I believe some longevity breakthrough will be the next big surprise.

Scarcity enables plot

You can see the theme across all these topics: Limited access, scarcity, sometimes secrecy. It serves stories well. Storytelling needs conflict, and science fiction offers a wealth of technology to chase after or hoard. At one end of storytelling this can be no more than a “MacGuffin” – it could be any shiny object everyone’s after. At the other end actual thoughtful examination of the technology can be done, but often takes the form of a plot revolving around a scarce resource nevertheless.

Science fictional storytelling prepares us for a world where advancements are expensive, scarce and withheld from the masses by an elite or a technological priesthood. When the real technologies arrive they’re made cheap, pervasive and impossible to get away from. The actual crisis in the real world comes from mass-saturation of a new technology.

I think sometimes the trope that science fiction is about the present, a thought experiment to see what already is, in a new light, is not quite the whole of science fiction. As technology develops more and more rapidly people really do have concerns specifically on that topic: What will happen next? Where is this all going? The “techno-thriller” subgenre is probably meeting that need most directly (though rarely adding any real insight, just plot).

To the extent readers really do want to engage with the question of how science and technology effect change, is the literature serving them well? I don’t think so. It’s primarily entertainment, but it serves as the foundation for most of our unconscious, casual thinking about the future.

The big failure of most classic and popular science fiction is one of economic and sociological imagination. The problems arising from the tech in the stories are solved as normal plot problems are: Directly confronting the issue, either with innovation or violence or both. It takes just a handful of characters, sometimes with a large army at their backs, but fundamentally a plan a handful of characters can concoct. The solution is one of engineering, treachery, or skill, not boring government policy implemented over decades.

  1. The Culture novels of Iain M. Banks are often cited. It’s a somewhat utopian and maybe surprising common favorite. Just below the surface is a dystopia where no humans truly make their own decisions, as the Minds – sentient god-like starships – vastly overpower them intellectually. Elon Musk has said they are his favorite vision of the future, and certainly he could do worse. He’s also mentioned Asimov’s Foundation, where a small group of “psycho-historians” guide civilization through a dark age, and Asimov’s robot stories with the “Three Laws of Robotics.”

    Mark Zuckerberg famously renamed his company “Meta” after the “Metaverse,” a VR + internet interface in Neal Stephenson’s “Snow Crash.” The Google founders also cite Stephenson.

    Almost every software and AI leader (Musk, Page, Brin, Hassabis, Altman, Amodei) steeped themselves in stories about singular, godlike AI “Minds” (Banks), secret scientific elites guiding humanity from above (Asimov’s Foundation), or rare, bespoke artifacts (Stephenson’s The Diamond Age). They spent years debating whether a singular AGI would wake up in a server rack and either murder us all or usher in post-scarcity space communism, maybe in part because of those books.

    Meanwhile, the one guy who explicitly refused to read science fiction, Jensen Huang (“I didn’t read those fucking books.”), didn’t treat AI as a mystical, unique god-mind from a novel. He treated it as mass-manufactured chips deployed to every data center on Earth to run ubiquitous software. He’s one individual who we may wish had read some of those books.

  2. Mega computers centralizing and planning society: Isaac Asimov’s “The Last Question.”

  3. “A Logic Named Joe” – Murray Leinster, “The Machine Stops” – E. M. Forster

  4. Apple IIc poster, William Gibson interview about the inspiration for Neuromancer – William Gibson

  5. Count Zero and Mona Lisa Overdrive – William Gibson

  6. The Simulacra – Philip K. Dick

  7. Even when classic and contemporary science fiction envisioned global communications networks, personal terminals, or virtual worlds, it routinely overlooked both mass-market commercialization and advertising as the economic engine:

    The Shockwave Rider – John Brunner: Envisions a continent-spanning data network and self-replicating computer “tapeworms,” but treats the network as an instrument of technocratic state surveillance rather than mass consumer entertainment and advertising.

    Ender’s Game – Orson Scott Card: Imagines the “nets” as high-minded public forums where two children using pseudonyms can sway global geopolitics through persuasive essays – with no ad networks, algorithms, or commercial spam in sight.

    Islands in the Net – Bruce Sterling: Depicts a world connected by “the Net,” yet the drama revolves around corporate diplomacy, data havens, and piracy rather than the Net itself being a mass-market advertising and attention machine.

    Earth – David Brin: Frequently credited with predicting the World Wide Web and ubiquitous citizen cameras, yet frames the global network around civic debate and public information rather than an attention economy funded by targeted ads.

    2001: A Space Odyssey – Arthur C. Clarke: Features flat-screen “newspads” plugged into a global information network and pocket “minisecs,” but imagines digital media as staid, traditional broadsheets rather than commercialized, ad-driven feeds.

    The Naked Sun – Isaac Asimov: Everyone on Solaria communicates exclusively through 3D holographic “viewing” rather than in-person contact – essentially a planet of shut-ins living online – yet it functions as a sterile utility for a tiny elite, completely devoid of mass media or advertising.

  8. “True Names” – Vernor Vinge

  9. Extended lifespan is usually present in classic and modern science fiction. It’s always hard to come by. Just a few examples more or less at random:

    The Immortals – James E. Gunn: The discovery of a handful of genetically immortal people has set off a craze in society to live as long as possible by any means necessary. For most this amounts to artificial organs, organ transplants, infusions and exotic treatments. Most resources go to medical care, while regular care remains out of reach for a large underclass. No one figures out how the mutation works (which would have changed everything) and instead they hunt down the immortals for their blood.

    Red Mars – Kim Stanley Robinson: Longevity is only for a small elite group, leading to political unrest.

    The Diamond Age – Neal Stephenson

    Dune – Frank Herbert

    Cyteen – C.J. Cherryh

    Beggars in Spain – Nancy Kress: The “sleepless” gene not only eliminates the need to sleep but grants slowed aging and is available only to a select few. While I don’t know if such a discovery would remain exclusive, if it did, Beggars in Spain is what the world would look like as a result.

    The Artificial Kid – Bruce Sterling: Bodies live for centuries, but minds need to be reset every few centuries to prevent insanity.

    Pandora’s Star – Peter F. Hamilton: The entire social hierarchy revolves around rejuvenation.

    Elizabeth Moon’s Serrano series: The elite live for centuries in exchange for staying out of government.

    A Deepness in the Sky – Vernor Vinge

    Andre Norton’s Galactic Empire / Central Control novels: The upper crust of society can live for centuries through some unspecified means.

  10. “Down and Out in the Magic Kingdom” by Cory Doctorow is an unusual contrasting work where everyone has basic needs met and immortality is achieved. The story explores what we’d do then. Social status is still in limited supply, so even roller coaster management becomes an epic drama. It’s a bit different from classic longevity as people survive death through cloning and brain uploading, a scenario I consider very implausible compared to simple medical life-extension. For the story it’s necessary to remove the fear of death.