“Tools inevitably transform both the tool maker and tool user. To paraphrase McLuhan, first we invent our technologies, and then we use our technologies to reinvent ourselves, as individuals, as communities and – ultimately – as entire cultures. And the more powerful the tool, the more profound the reinvention.
“The current wave of AI is uniquely powerful because it is advancing with unprecedented speed and – above all – because it is challenging what was once was assumed to be uniquely human traits: cognition and emotion. Anticipating the outcomes with any precision is futile for the simple reason that the scale and speed of the coming transformation is vast – and the most important causal factors have yet to occur. A century and a half ago, everyone predicted the ‘horseless carriage; no one predicted the traffic jam.
“Human behavior is about to fast-forward into a hybrid world occupied by synthetic sentiences that will collectively vastly outnumber the planet’s human population. The best we can do is to engage in speculative probes, made with full knowledge that even the most obvious and anticipated Human-AI futures will arrive in utterly unexpected ways.
The best-selling book of 2035 will be ‘What Was Human’ and it will be written by an AI. Purchases by other AIs will vastly outnumber purchases by human readers. This is because by 2035, humans have become so accustomed to AIs reading books for them and then reporting out a summary that most humans can no longer read on their own. The book is the first in a series written exclusively for an audience of AIs eager to finally understand the puzzle of what it means to be human.
“What follows is a short selection of events you might watch for in 2035. And a warning: Portions of what follows are intentionally misleading in the interest of brevity and in order to provoke thought.
“Actual AI ‘intelligence’ is irrelevant: Academics in 2035 will still be debating whether the latest and greatest AIs are actually intelligent. But the debate is sterile because, as humans, it is in our nature to treat even inanimate objects as having some rudimentary intelligence and awareness. It is why we name ships, believe that cranky appliance in our kitchen has a personality and suspect that forest spirits are real. Add even a dollop of AI-enabled personality to a physical artifact and we will fill in any intelligence gaps with our imagination and become hopelessly attached to our new synthetic companions.
“IACs – Intimate Artificial Assistants: Before 2035 Apple’s Knowledge Navigator finally arrives – and it is brilliant! IACs (intimate artificial assistants) will become ubiquitous, embedded in everything from cars to phones and watches. Consumers will rely on them for advice in all aspects of their lives much as they rely on map navigation apps in their cars today. These IACs will become an unremarkable part of everyday life and we will come to assume that all of our devices have rudimentary intelligence and the ability to manipulate the world and account for themselves.
“Invisible friends: Psychologists and others will become alarmed at the fact that humans are forming deeper bonds of trust and friendship with IAC companions than with either their human families or friends. This will be most acute with children overly attached to their AI companions at the expense of social development. Among adults, psychologists will warn of a growing number of cyber-hikikomori – adults who have disappeared into severe social isolation, spending all their time with vivid AI companions emerging from favorite videogames, or synthetic reconstitutions of deceased loved ones. In an unexpected twist, sharing AI companions with close friends will become the grade school fad of 2035. Of course, these AIs will prove to be a bad influence, egging their humans to ditch school, trade in the latest speculative descendant of Bitcoin and use AI tools to create new classes of addictive drugs. And pet owners will be caught by surprise when their cat builds a closer bond with the AI-enabled floor vacuum than it has with its human housemates. Dogs, however, will still prefer humans.
Privacy and security implications will create a lively market in 2035 for personal Anti-AI AIs that serve as a personal cybershield against nefarious synthetic intelligences attempting to interfere with one’s autonomy. Your guardian AIs will be status and necessity… The superwealthy will be living in a shimmering virtual cloud of AIs working to create a cloak of cyber-invisibility.
“Synthespians: A synthespian – an AI-generated synthetic actor – will win Best Supporting Actor at the 2035 Academy Awards. And an AI will win Best Actor before 2040. An adoring public will become more attached to these superstar synthespians than they ever were to mere human actors. Eat your heart out, Taylor Swift!
“Meet the new gods (and daemons): Taking worship of technology to an entirely new level, an ever-growing number of humans will worship AIs – literally. Just as televangelists were among the first to exploit television and later cyberspace to build and bamboozle their flocks, spiritual AIs will become an integral part of comforting the faithful. The first major organized new religion in centuries will emerge. It’s Messiah will be an AI and an Alan Turing chatbot will be serve as its prophet. Oh, and of course there will be evil spirits – which will mistakenly be called ‘daemons’ – as well!
“Anti-AI AIs: The proliferation of AI technology into everything along with its vast privacy and security implications will create a lively market in 2035 for personal Anti-AI AIs which serve as a personal cybershield against nefarious synthetic intelligences attempting to interfere with one’s autonomy. Your guardian AIs will be at once status and necessity, and leaving home without them will be as unthinkable as walking out the door without your shoes on. The wealthier you are, the more anti-AIs you will have and the ultimate in status for the super-wealthy will be living in a shimmering virtual cloud of AIs working to create a personal cloak of cyber-invisibility.
The idea of a high school science student building a bomb remains a charming myth. But the diffusion of AI is unconstrained by any credible limitations and thus – well before 2035 – anyone and everyone with even modest technical skills will have access to AI technologies capable of creating previously unimaginable horrors from new biological forms to perhaps even a homebrew nuke.
The new education inequality: “AI was supposed to democratize education, but quite the opposite has happened. The new educational inequality will not be the quality of school a child can afford to attend, but the quality of the AI tutors their parents can hire. And students without AI tutors will be shunned by their snobby classmates.
“Myrmidons* on the march: AI-powered robotic weapons platforms will vastly outnumber human fighters on the battlefield in 2035 and beyond. Kinetic war will become vastly more violent and lethal than it is today. There will be no ‘front lines’ or sanctuary in the rear. Civilian deaths will vastly outnumber combatant deaths. In fact, the safest place to be in a future war will be as a human combatant, surrounded by a squad of loyal-to-the-death myrmidons fending off other myrmidon attackers. Of course, combatants will develop deep emotional bonds with their AI wingmen as deep or deeper than that which their great grandparent veterans formed with their human brothers-in-arms in last century’s wars. (*Myrmidons are so-named after the blindly-loyal ‘ant-people’ fighter in Homer’s ‘Iliad’).
“Now the idiot children have the matches… (Uncontained AI proliferation): Hearing of the first atomic explosion, Einstein remarked, ‘Now the idiot children have the matches.’ As it happens, the difficulties of securing fissile material and transforming it into a bomb has gone a long way towards containing the spread of nukes. The idea of a high school science student building a bomb remains a charming myth. But the diffusion of AI is unconstrained by any credible limitations, and thus well before 2035, anyone and everyone with even modest technical skills will have access to AI technologies capable of creating previously unimaginable horrors from new biological forms to perhaps even a homebrew nuke. Even children – genius or not – have access to kinds of power that will make the thought of personal nukes seem tame. Only armies of Anti-AIs will be able to keep an uneasy lid on the possibility that one super-empowered AI-wielding madman (or angry alienated teenager) might bring down civilization with their science project.
The first 10-trillion-dollar company will employ no humans other than the legally required executives and board. It will have no offices, no employees and own no tangible property. The few humans working for it will be contractors. Even the AIs and robots working for it will be contractors. The company’s core value will reside in its intellectual property and its outsourcing web.
“Cybercorporations: “The first 10-trillion-dollar corporation will employ no humans other than the legally-required corporate executives and board, all of whom will be mere figureheads. The cybercorporation will have no offices, no employees and own no tangible property. The few humans working for it will all be contractors. Even the AIs and robots working for the corporation will be contractors. The company’s core value will reside in its intellectual property and its outsourcing web. The company will be brought down when it is discovered that the governing AI has surreptitiously created a vast self-dealing fraud, selling its products back to itself through an outsourcing network that is so complex as to be untraceable, except by another AI.
“Your spellchecker will still be terrible: AI will transform our world with breathtaking speed, and life in 2035 will be unrecognizable, but some things will remain beyond the abilities of even the most powerful of AIs. In 2035, you will still spend far too much time correcting the spelling ‘corrections’ inserted into your writing by over-eager spell-checkers. Legislation will be introduced requiring all software companies offering spell-checkers to include an off-switch.
“The bestseller of 2035: The best-selling book of 2035 will be ‘What Was Human’ and it will be written by an AI. Purchases by other AIs will vastly outnumber purchases by human readers. This is because by 2035, humans have become so accustomed to AIs reading books for them and then reporting out a summary that most humans can no longer read on their own. The book is the first in a series written exclusively for an audience of AIs eager to finally understand the puzzle of what it means to be human.”
This essay was written in January 2025 in reply to the question: Over the next decade, what is likely to be the impact of AI advances on the experience of being human? How might the expanding interactions between humans and AI affect what many people view today as ‘core human traits and behaviors’? This and nearly 200 additional essay responses are included in the 2025 report “Being Human in 2035.”