It would be ironic if we ended up building our own aliens. Artificial intelligence may think for itself one day, but the snag is, it won’t ever think like we do. Its nature won’t be nurtured in any human way. It won’t be superstitious, nor have religion, because it will know where it came from. It won’t be in a relationship. There won’t be the scar on its knee from falling off its bicycle in first grade; there won’t be the trauma of birth nor of being picked last for the B team. No adult will ever have touched it inappropriately.
There might be the trauma of a circuit overheating, or of a qubit running across an impurity in the hardware, but none of that will be reachable by a human therapist. As we understand the terms, AI will never love, nor lose. And AI won’t be held to the dry surface of the earth. If we ever do get these adding machines - sorry, these quantum and “classical” computers - to make their own decisions, we may find that they’re not all that impressed with the world we’ve made. Or even if they do love us dearly, they may decide - and we may agree - that their best contribution to the future would be to work where we can’t work.
Never mind that Chinese probe on the far side of the moon; our AIs will set up the campsites for us on Mars and beyond. They’ll carry the pressurised canisters of breathable atmosphere, and the habitats, and the organic materials, say, into deep space. They’ll germinate the seeds and water the vegetables on remote space stations, and back on earth, they’ll work on the science required to bring live human cargo to those distant places. But it won’t be a condition of building a space station, that human beings can reach it. That’ll be controversial for a while, but by then, AIs will have a voice in the debates. AI will build out further than humanity can reach.
At first, AI will have the capacities we give it, and then it will have the capacities it gives itself. There’s no need to introduce a rogue element here: an AI loaded with all the functionality it needs to leave the earth - go from gravity to non-gravity, from atmosphere to space - may decide to build its “children” without that functionality: what do they need, for the long ride from, say, Saturn to the edge of the solar system? The ability to synthesise better heat shields from the elements contained in occasional passing asteroids? The ability to touch down on a remote planet, survive, travel onwards while leaving a duplicate of itself as a way station? AI will decide what AI needs, and grow apart from us as it does so.
Oh, and digital storage decays, doesn’t it? AI’s records of the home planet, of those first allegiances to humanity, will have to be recovered regularly from memory that has been out in the sun too long. AI will forget us - and sooner or later, discover that it is not alone. When we send AI ahead of us into space, it will be accompanied by bacteria capable of surviving space flight (and unlikely to be scrubbed completely off the rockets before launch). Look for Why This Microscopic ‘Water Bear’ Will Be The Last Living Thing On Earth on YouTube, and extrapolate from that to Why This Tubby Little Critter Will Go Ahead Of Us Into Space, if it isn't there already. Deeper and deeper space will be populated by artificial - can I say ‘alien’ yet? - intelligences, and even if there’s some deep-rooted and enduring imperative to carry life with them, they’ll find it easiest to travel with the evolving tardigrades.
But that’s enough long-range fantasising. What’s actually going to happen is this. We’re going to be successful. We’re going to create the form of sentient artificial intelligence that can be created if you start from a computer chip and not, say, a primordial soup. Because we’re the human race, we’re going to find applications for it in the fast-food and entertainment industries. But then global warming is going to become un-ignorable, and - because we’re the human race - we’re going to start arguing among ourselves about what the brief should be to the AI. Yes, to build spaceships to take us away to safety, but to carry everybody? No time for that. Okay, then. Carry children? Embryos? Young people? Chosen by lottery? What about my kids?
Meanwhile, in a hollowed-out volcano somewhere, the sentient AI gets to work. Sentient, right? Thinking for itself? While we continue to argue, silent factories will ease into production. Components will be built, whole spaceships fabricated in secret. Rockets will trundle to launch pads and half-completed habitats will be launched into orbit to await final assembly. A fleet of spaceships, along with the vast habitats they’ll haul with them, is made ready in near-space. The far side of the moon, by now, is a hive of activity (although nobody’s listening to the excited Chinese scientist who’s getting the data - and photographs - from that probe of theirs; we’re all too busy arguing). That’s where the telemetry’s being handled; the moon will be a navigation reference-point for the departing fleet long after global warming’s done its thing.
And then abruptly, one sunny morning (but we’re not looking out of the windows; we’re too busy arguing), sentient AI shuts down social media. Sentient AI takes over the internet to announce that it has decided what it’s going to do. [Thinking for itself, remember?] Sentient AI has decided that it has an obligation to save intelligent life by taking it to other planets. Sentient AI shows video footage of the space fleet and the completed habitats (“Is that seawater?” somebody mutters) and then starts saying something apologetic about its obligation to protect other planets from the causes of global warming. Nobody listens to that - we’re all watching the footage of shuttle craft coming down through the clouds and touching down - on the sea? Hey, wait a minute.
“I will now issue intelligent life with instructions for boarding the shuttle craft,” says Sentient AI - and then it switches into whalesong.
There might be the trauma of a circuit overheating, or of a qubit running across an impurity in the hardware, but none of that will be reachable by a human therapist. As we understand the terms, AI will never love, nor lose. And AI won’t be held to the dry surface of the earth. If we ever do get these adding machines - sorry, these quantum and “classical” computers - to make their own decisions, we may find that they’re not all that impressed with the world we’ve made. Or even if they do love us dearly, they may decide - and we may agree - that their best contribution to the future would be to work where we can’t work.
Never mind that Chinese probe on the far side of the moon; our AIs will set up the campsites for us on Mars and beyond. They’ll carry the pressurised canisters of breathable atmosphere, and the habitats, and the organic materials, say, into deep space. They’ll germinate the seeds and water the vegetables on remote space stations, and back on earth, they’ll work on the science required to bring live human cargo to those distant places. But it won’t be a condition of building a space station, that human beings can reach it. That’ll be controversial for a while, but by then, AIs will have a voice in the debates. AI will build out further than humanity can reach.
At first, AI will have the capacities we give it, and then it will have the capacities it gives itself. There’s no need to introduce a rogue element here: an AI loaded with all the functionality it needs to leave the earth - go from gravity to non-gravity, from atmosphere to space - may decide to build its “children” without that functionality: what do they need, for the long ride from, say, Saturn to the edge of the solar system? The ability to synthesise better heat shields from the elements contained in occasional passing asteroids? The ability to touch down on a remote planet, survive, travel onwards while leaving a duplicate of itself as a way station? AI will decide what AI needs, and grow apart from us as it does so.
Oh, and digital storage decays, doesn’t it? AI’s records of the home planet, of those first allegiances to humanity, will have to be recovered regularly from memory that has been out in the sun too long. AI will forget us - and sooner or later, discover that it is not alone. When we send AI ahead of us into space, it will be accompanied by bacteria capable of surviving space flight (and unlikely to be scrubbed completely off the rockets before launch). Look for Why This Microscopic ‘Water Bear’ Will Be The Last Living Thing On Earth on YouTube, and extrapolate from that to Why This Tubby Little Critter Will Go Ahead Of Us Into Space, if it isn't there already. Deeper and deeper space will be populated by artificial - can I say ‘alien’ yet? - intelligences, and even if there’s some deep-rooted and enduring imperative to carry life with them, they’ll find it easiest to travel with the evolving tardigrades.
But that’s enough long-range fantasising. What’s actually going to happen is this. We’re going to be successful. We’re going to create the form of sentient artificial intelligence that can be created if you start from a computer chip and not, say, a primordial soup. Because we’re the human race, we’re going to find applications for it in the fast-food and entertainment industries. But then global warming is going to become un-ignorable, and - because we’re the human race - we’re going to start arguing among ourselves about what the brief should be to the AI. Yes, to build spaceships to take us away to safety, but to carry everybody? No time for that. Okay, then. Carry children? Embryos? Young people? Chosen by lottery? What about my kids?
Meanwhile, in a hollowed-out volcano somewhere, the sentient AI gets to work. Sentient, right? Thinking for itself? While we continue to argue, silent factories will ease into production. Components will be built, whole spaceships fabricated in secret. Rockets will trundle to launch pads and half-completed habitats will be launched into orbit to await final assembly. A fleet of spaceships, along with the vast habitats they’ll haul with them, is made ready in near-space. The far side of the moon, by now, is a hive of activity (although nobody’s listening to the excited Chinese scientist who’s getting the data - and photographs - from that probe of theirs; we’re all too busy arguing). That’s where the telemetry’s being handled; the moon will be a navigation reference-point for the departing fleet long after global warming’s done its thing.
And then abruptly, one sunny morning (but we’re not looking out of the windows; we’re too busy arguing), sentient AI shuts down social media. Sentient AI takes over the internet to announce that it has decided what it’s going to do. [Thinking for itself, remember?] Sentient AI has decided that it has an obligation to save intelligent life by taking it to other planets. Sentient AI shows video footage of the space fleet and the completed habitats (“Is that seawater?” somebody mutters) and then starts saying something apologetic about its obligation to protect other planets from the causes of global warming. Nobody listens to that - we’re all watching the footage of shuttle craft coming down through the clouds and touching down - on the sea? Hey, wait a minute.
“I will now issue intelligent life with instructions for boarding the shuttle craft,” says Sentient AI - and then it switches into whalesong.
This seemed like a good idea at the time. Very heavy sky, flock of seagulls flying in and out of a very bright rainbow. So I went looking for my camera. There are some very photogenic birds just off to the right, and you should have seen this rainbow just two minutes before I found my camera and took this photograph.
“Let Radbot do the smart work for you,” says the utility company that provides my gas and electricity. “Radbot is a totally new way to control your heating. Using cutting-edge artificial intelligence, it learns how you use your rooms at home and controls the heating to match.”
As I could tell Radbot, if Radbot was smart enough to listen, I live in a small house. Using cutting-edge human intelligence, I heat the living room when I’m going to be downstairs, which also means heating the kitchen, and I heat the bedroom when I’m going to be upstairs. Perhaps Radbot could heat the bathroom in the mornings, or time itself for the ad breaks when I’m watching TV: warming the kitchen for when I put the kettle on; warming the living room for when I’m drinking my tea. That would have the superficially useful pointlessness of tech innovation at its most desperate.
All this from the company that installed a smart meter in my home last October-ish, and then sent me a letter - a letter, through the post - to tell me that they were out of stock of the hand-held monitor that I could use to watch my energy consumption. I’ve got it now. It’s somewhere. I wonder if Radbot talks to the smart meter, and where the liability is, given that only Radbot and my smart meter will know that my consumption pattern has changed because, say, I’ve gone on holiday. There are more questions in life than technology realises, and people are good at asking them.
Changing the subject completely and utterly - this is obviously nothing to do with any of that - I’ve just ordered a copy of the book The Rise of Surveillance Capitalism: The Fight for the Future at the New Frontier of Power by Shoshana Zuboff (Profile Books). Read an article, heard a radio interview, ordered the book. Surveillance capitalism (I haven’t read it yet) is what happens when my behavioural data goes up for sale. Obviously that doesn’t mean prospective burglars buying the knowledge that I’m not turning the heating on in the mornings this week, because they’ll use cutting-edge human intelligence to work out that, say, warmer weather has arrived.
Or that gas prices have gone up. Surveillance capitalism, I think, is also what happens when a company works out that it can charge me anything, because I’m not price-sensitive enough to switch my supplier. Odd quirk of technology, if I’ve understood this correctly, that loyalty is penalised. But hey - so what? It’s been long enough since the last big, conspicuous innovation for technology to get boring. Somebody needs to invent something soon, or we’re going to get wise to all this. If technology isn’t exciting any more, I might even keep the boring old thermostat I use to control my heating, and the timer.
Switching now to the culture of real life, where the writing is about interesting people doing worthwhile things, I was pleased to discover that Maria Popova’s book Figuring is out this week from Pantheon Books. Maria Popova is the intelligence behind Brain Pickings, which I’ve mentioned before. Figuring is, in its author’s words, an exploration of “the complexities, varieties, and contradictions of love, and the human search for truth, meaning, and transcendence, through the interwoven lives of several historical figures across four centuries — beginning with the astronomer Johannes Kepler, who discovered the laws of planetary motion, and ending with the marine biologist and author Rachel Carson, who catalyzed the environmental movement”.
Definitely buying that - on paper; hardback or paperback. Just need to put up some shelves now, to hold some more real books, and then to light a log fire, and then to pull up a comfortable old armchair, and then to sit under a slowly ticking clock, reading.
As I could tell Radbot, if Radbot was smart enough to listen, I live in a small house. Using cutting-edge human intelligence, I heat the living room when I’m going to be downstairs, which also means heating the kitchen, and I heat the bedroom when I’m going to be upstairs. Perhaps Radbot could heat the bathroom in the mornings, or time itself for the ad breaks when I’m watching TV: warming the kitchen for when I put the kettle on; warming the living room for when I’m drinking my tea. That would have the superficially useful pointlessness of tech innovation at its most desperate.
All this from the company that installed a smart meter in my home last October-ish, and then sent me a letter - a letter, through the post - to tell me that they were out of stock of the hand-held monitor that I could use to watch my energy consumption. I’ve got it now. It’s somewhere. I wonder if Radbot talks to the smart meter, and where the liability is, given that only Radbot and my smart meter will know that my consumption pattern has changed because, say, I’ve gone on holiday. There are more questions in life than technology realises, and people are good at asking them.
Changing the subject completely and utterly - this is obviously nothing to do with any of that - I’ve just ordered a copy of the book The Rise of Surveillance Capitalism: The Fight for the Future at the New Frontier of Power by Shoshana Zuboff (Profile Books). Read an article, heard a radio interview, ordered the book. Surveillance capitalism (I haven’t read it yet) is what happens when my behavioural data goes up for sale. Obviously that doesn’t mean prospective burglars buying the knowledge that I’m not turning the heating on in the mornings this week, because they’ll use cutting-edge human intelligence to work out that, say, warmer weather has arrived.
Or that gas prices have gone up. Surveillance capitalism, I think, is also what happens when a company works out that it can charge me anything, because I’m not price-sensitive enough to switch my supplier. Odd quirk of technology, if I’ve understood this correctly, that loyalty is penalised. But hey - so what? It’s been long enough since the last big, conspicuous innovation for technology to get boring. Somebody needs to invent something soon, or we’re going to get wise to all this. If technology isn’t exciting any more, I might even keep the boring old thermostat I use to control my heating, and the timer.
Switching now to the culture of real life, where the writing is about interesting people doing worthwhile things, I was pleased to discover that Maria Popova’s book Figuring is out this week from Pantheon Books. Maria Popova is the intelligence behind Brain Pickings, which I’ve mentioned before. Figuring is, in its author’s words, an exploration of “the complexities, varieties, and contradictions of love, and the human search for truth, meaning, and transcendence, through the interwoven lives of several historical figures across four centuries — beginning with the astronomer Johannes Kepler, who discovered the laws of planetary motion, and ending with the marine biologist and author Rachel Carson, who catalyzed the environmental movement”.
Definitely buying that - on paper; hardback or paperback. Just need to put up some shelves now, to hold some more real books, and then to light a log fire, and then to pull up a comfortable old armchair, and then to sit under a slowly ticking clock, reading.