Reality Isn’t Sci-Fi—It’s More Startling
I’m just back from the 80th World Science Fiction Convention, held in my native city of Chicago this year. A delightful return to normalcy, if you discount the masks, vaccination requirements, and online-only participants. And it’s brought to mind comments, from long ago, by two different authors about the tension between sci-fi tech and real-life tech.
I’ll start with Vernor Vinge, whose most famous novel is A Fire Upon the Deep. His 1986 book, Across Realtime, is set in part in the 23rd century. In the afterword, Vinge apologizes that the technology in his book, set two centuries in the future, is not advanced enough! He explains that a more accurate extrapolation, based on current trends, would render any story set two centuries from now almost incomprehensible.
The denizens of that time have fascinating weapons, including pistols that can fire autonomous bullets that seek their targets. In 1986, that capability was absolutely far-fetched. But no longer. We now have precision-guided missiles; the beginnings of laser weapons that use light itself as its armament; semi-autonomous armed drones. You can purchase a sniper rifle (for $9,000) that fires the shot for you: put the red dot on the target, depress the trigger, and the rifle will fire when it determines conditions are optimal to hit the target. Technology that was once considered futuristic, likely impossible, technically too far advanced, is now available.
Thirty years ago, when testing some of the best ASR available, I could barely get the system to recognize a seven-digit phone number. Today my wristwatch accepts voice input commands. The second most amazing development over the past 30 years is the public’s ability to use speech interfaces to access the consumer empires of Big Tech’s U.S. corporations, without the need for training to a user’s specific voice. The first most amazing is applications that translate languages.
So let’s try a modified version of Moore’s law. Moore’s law states that the computational power of computer chips increases twofold every 18 months. Let’s assume that speech technology and artificial intelligence doubles in capability, accuracy, and (in the case of artificial intelligence) scope every seven years. Twenty years from now, what can we expect from these technologies—eight times more advanced than what’s available today?
Returning to science fiction, I recall an exchange with the late, great Jack Vance, still my favorite author. At the 1992 Worldcon I asked him a simple question: “Why don’t you have robots in your stories?” Vance’s stories had a singular lack of almost any computer interactions. His science fiction was filled with other futuristic ideas: space travel between the stars; aliens, perhaps the best ever described, with truly alien ways of thinking; polities that spanned wide swaths of the galaxy. Vance replied, in part, “I don’t think robots are interesting.” He preferred to dwell on human society and the infinite possibilities of life.
To combine the viewpoints of Vance and Vinge, let’s discuss what to expect in 20 years, from the human vantage, when threefold faster and better speech technology combines with—I deliberately do not use the adjective “better”—threefold more advanced AI.
First, the dystopia. China, assuming that the current rulers manage to keep their Communist government creaking along, will use that threefold increase to further augment its “social credit” system. Spoken conversations in any public place, or using any form of telecommunications, will be monitored. Anonymity will be nearly impossible.
Sound bad? The United States will not be far behind. As the cost of cellular service continues to decline, service providers may find it more profitable to offer free service—if the customer assents to anonymous monitoring of calls, extraction of relevant speech data, and sale of that data to advertisers, political parties, and other entities. Starting with “sensitive” public places, such as airports, expect constant monitoring of all in-person conversations. The list of places will expand (e.g., after a mass shooting) to include schools, shopping malls, and eventually entire neighborhoods.
But with the possibility of bad comes the possibility of good. If I can purchase and manage my own systems, and keep them secure, then I will finally have something I’ve coveted for years: a personal gentleman’s gentleman. Do I want tickets to a movie? A reminder of an upcoming appointment? A bit of research for my next article? Perhaps I need a new shirt, and I need to know what’s both fashionable and consistent with my wardrobe. All these are (likely) available today via various services, but an integrated service under my own control with integrated, expansive knowledge would truly be delightful.
Moshe Yudkowsky, Ph.D., is president of Disaggregate Consulting and author of The Pebble and the Avalanche: How Taking Things Apart Creates Revolutions. He can be reached at firstname.lastname@example.org.