A few years ago, my buddy, Robert Sawyer postulated that because we now use computers as a critical tool for research, Moore’s Law applies to scientific accomplishment as well.
He started with a simple postulate—assume that in the first decade of the 21st century we have already accomplished as much scientific advancement as we accomplished in the entire 20th century—amazing discoveries in astronomy, paleontology, materials, medicine, robotics, etc.
Now, let’s try a thought experiment. If we apply Moore’s law and assume that the rate of scientific advancement doubles at the same rate as the computer power that we apply to research, then we can project that we will likely accomplish a whole 20th century’s worth of scientific advancement in 5 years—by 2015. As the rate continues to double, we’ll accomplish a century’s work in 2.5 years, then 1.25 years, 7.5 months, 3 months and 3 weeks, then a smidge less than two months, one month, two weeks, one week, then 3.5 days, 1.75 days, and if you ignore Zeno’s paradox, by the end of 2020 we will be accomplishing a century’s worth of research every day, and two weeks later, every second. And after that…?
Will that be when The Singularity occurs?
In math, a singularity is a point where a function demonstrates extreme behavior. The Singularity , as defined by Vernor Vinge and Ray Kurzweil, will occur with the technological creation of superintelligence. Such a world may be impossible to predict because us poor present-day humans are unable to comprehend what superintelligent entities will want or how they’ll behave to achieve their goals. (Well, yeah, okay—but life has one fundamental rule: survive. Start with that and everything else follows.)
It may be that The Singularity is nothing more than a technological ‘rapture’ — an event of some interest to those who believe in it, but not necessarily one that the rest of us are expecting. In October of 1951, The Magazine of Fantasy Science Fiction published a story by Richard Deming called “The Shape Of Things That Came.” In that story, a young reporter uses his uncle’s time-nightshirt (what a silly idea, he should have used a time-belt) to travel from 1900 to 1950. When he returns, he writes about what he has seen — highways and cities full of cars, huge airplanes traveling coast-to-coast, skyscrapers sixty and eighty and a hundred stories tall, telephones everywhere, radio broadcasting, moving pictures with color and sound, television beaming into every home—but his editor rejects the tale because of its essential implausibility. Paraphrasing: “Yes, all of those things are certainly possible at some point in the distant future—but not in fifty years. What is impossible to believe is the timespan. Many of the people in your tale are already born. The human mind simply cannot deal with so much change in a single lifetime.”
In the fifty years since that story was first published, we’ve seen even more astonishing changes in our science and technology: nuclear power, organ transplants, multiple trips to the moon, solar panels, communication satellites, space probes, an international space station, supersonic jets, genetically modified crops, digital information technology, the widespread use of lasers for transmitting and storing information (as well as for teasing cats), globally-connected cell phones, personal computers of all sizes and vast libraries of applications, the incredible reach and versatility of the internet, video games (of course), Viagra, and so much more.
So maybe, just maybe, when and if The Singularity occurs, it will be just one more thing that human beings take in stride—and then complain about because that’s one of the things that machines still can’t do. For most of us, technological advances are a way to address the fundamental laziness of the species—we’re looking for an easier way to get the job done. Good, fast, cheap, we’re happy with any two out of three.
We might take a clue from what happened last century. The two inventions that had the most impact on the 21st century came in the second half—the microchip and the laser. To a great degree they were unexpected and mostly unpredicted. The laser was a lab curiosity for decades. And even after the first microchips were fabricated, industry still didn’t recognize the true potential—not until a couple guys in a garage showed them what a microchip could really do.
It’s very likely that today’s lab curiosities represent possibilities that will redesign our world. Here are a few things to watch out for. (We’ll check back in a few years and see if my crystal ball needs recalibrating.)
First there were bucky-balls, then bucky-tubes, now unrolled bucky-tubes in flat sheets, only an atom thick. Already being touted as the miracle material of the future, graphene is still just a lab curiosity because nobody knows how to manufacture it in industrial quantities, but if graphene could be manufactured efficiently, it would be the plastic of the 21st century. A lot of people believe the problem is solvable.
Researchers at IBM have already demonstrated high-speed circuits on a graphene substrate . What happens when we move from gigaherz processors to teraherz processors? Yes, everything we do now will be faster, effectively instantaneous, but just as the gigaherz CPU made speech recognition and photo-editing and video processing practical what other labor-intensive tasks will the teraherz CPU be able to handle without breaking into a sweat? Add parallel processing to that and we’re talking hellaflops.
But more than that, graphene has incredible physical strength. Researchers at Columbia University have proven that graphene is the strongest material ever measured, some 200 times stronger than structural steel. Quote: “It would take an elephant, balanced on a pencil, to break through a sheet of graphene the thickness of saran wrap.” Other scientists have layered multiple graphene sheets into a paper-like form that is six times lighter than steel, two times harder, has 10 times higher tensile strength and 13 times higher bending rigidity.
Perhaps someday, layered graphene will be used in cars, planes, trains, buses, spacecraft, robots, perhaps even buildings. The weight savings alone will provide significant fuel economy and the increased strength will give greater structural integrity and safety. It could also show up in military armor. We might see it used in lightweight patio furniture or rugged laptop shells or even as rollbars in some future generation of hybrids.
The predictions for graphene have not yet been tested by reality—but if graphene really is a miracle material, it will have enormous impact on the global infrastructure.
The orbital elevator—the “beanstalk”—has become a popular science fiction idea, first showing up in novels by Arthur C. Clarke and Charles Sheffield, but also appearing in books by other authors as well ( The Fountains Of Paradise by Arthur C. Clarke, The Web Between The Worlds by Charles Sheffield, Friday by Robert A. Heinlein, Red Mars by Kim Stanley Robinson, and Jumping Off The Planet by David Gerrold). Drop a cable from orbit and run elevators up and down, reducing the cost of orbital insertion by at least an order of magnitude. It may be that graphene cables will be the miracle material that lets us build one, but an orbital elevator will require a cable 40,000 miles long, almost enough to wrap around the Earth twice and that requires manufacturing on a scale never before attempted. And right now, graphene is still a long way from ‘proof-of-concept.’
Considering the cost of boosting even a single pound into orbit, such a cable will have to be manufactured in space and that means the factory to build it will also have to be built in space. At the moment, we can’t afford to lift that much weight out of the gravity well. It could be a trillion dollar investment. And the recovery of that cost could take generations. The economics of an orbital elevator, as well as the physics, are enormous challenges. Overall, the sheer outlandishness of the idea may be one of the reasons why it hasn’t captured the public imagination, so it may be that launch catapults (or some other technology) will be more cost-effective in the meantime. A practical beanstalk doesn’t seem likely in the foreseeable future—but remember that as late as 1960, most futurists (science fiction writers) still thought that the first moon landing wouldn’t occur until sometime in the 90’s, so maybe we could be similarly surprised.
Much more likely, the first uses of super-cables in space will probably be tethered satellites or even whirling bolos slinging vehicles and probes out toward the other planets, but the real impact will be here on Earth long before that.
Research into super-cables is going to produce some surprising industrial uses groundside, like a super-long suspension bridge across the strait of Gibralter, or perhaps as unbreakable tethers for energy-generating jet-stream kites, and certainly new possibilities in architecture—like holding up a super-tent of graphene fabric to create a gigantic weather-proof facility. Changing the properties of any single element in the industrial equation will create engineering possibilities that are not immediately foreseeable, but always look inevitable after the fact.
Robots: Already hanging with pop stars.
Robots are an easy prediction. Karel Capek created the word “robot” in a 1921 play, “Rossum’s Universal Robots,” and a lot of other science fiction writers began playing with the idea almost immediately, most notably Isaac Asimov. Robots have been a familiar fixture in a lot of science fiction movies, sometimes as good guys, sometimes not. Engineers were pondering the mechanics of robotics long before Walt Disney put an animatronic Lincoln on display for the 1964-65 World’s Fair in New York, but it wasn’t until brains, muscles, and power-supplies became small enough and efficient enough that we could begin to project an evolutionary timeline. A quick rummage through YouTube demonstrates that all the necessary pieces are finally falling into place.
One company is demonstrating a robot that can walk and even run, another shows a robotic face with a wide variety of expressions, a third displays a robot that can pick up and manipulate objects, catch balls and juggle them. Still other companies are working in intelligence engines that can comprehend complex language tasks. Nuance already sells pretty good speech recognition technology and Google has a car that can drive itself. IBM has a computer that can win at Jeopardy . And beyond that, a lot of other companies are working to develop smaller, more efficient motors and improved battery technology. All of these pieces are the synergistic parts of a much larger whole.
What the end-product will look like, however, is still a work in progress. We can imagine robots being put to work in the house, in business, in construction, in entertainment, in rescue operations, and certainly for military applications as well. But the first humanoid robots are likely to be simple, stupid, and disappointing—they’re also going to be expensive. People will see them as a good idea, but unable to live up to promises and expectations. Vista on legs.
Nevertheless, robots are inevitable. The first widespread use of robots will be in theme parks. Disneyland and Universal will use robots to portray creatures like dinosaurs and giants and dwarves. You can expect to see robot dancers in music videos too, but the real breakthrough will occur when robots start taking on more mundane tasks. We’ll see them as bartenders or aides for the sick and elderly. Robots could be put to work in hotels—you try changing sheets, lifting twenty or thirty mattresses a day. At the point a robot is cheaper than hiring a human, it’s inevitable. The job market will change when whole classes of human workers could become redundant.
And don’t forget Gerrold’s umpteenth law: Whatever technology humans invent, humans will also find a way to use that same technology for sex. So robotic sex partners are also inevitable—in brothels, for overnight rental, or even for purchase. (There was a young man from Racine, who invented a screwing machine. Concave or convex, it could serve either sex, entertaining itself in-between.) Because robots don’t get headaches. It is also possible that eventually, we will use robots as real-world avatars—surrogates—sending them out into the world to run errands for us, with remote control available where necessary.
Where robots will likely demonstrate their most critical value will be in hazardous situations. Robots will be used for military reconnaissance, for defusing bombs, and perhaps eventually even for assault duties. Robots will certainly be used for handling dangerous materials and toxic waste cleanup. And I can even imagine robots patiently and methodically cleaning up oil spills—even rescuing and cleaning seals and seabirds.
The first robots are likely to start showing up before this decade is over. (Perhaps Apple will market the iRobot.) Once the initial sugar-rush wears off, that’s when we’ll start finding out what we really want robots to do for us.
Starting in the early fifties, futurists started telling us that flying cars and flat-screen televisions were only ten years away. We finally got affordable flat-screen TVs in 2005. We still don’t have flying cars , and it is unlikely that we ever will.
First of all, a flying car is not cost-effective. Have you checked gas prices lately? A flying vehicle has to do a lot of work just to stay aloft. What kind of gas mileage are you going to get with a flying car?
Second? It’s impractical . Do you have room in your driveway for a landing pad? And is there a landing pad at your intended destination? Where do you need to go that demands a flying car? The supermarket? Picking up the kids from school? Any trip less than thirty or forty miles is probably going to be more trouble than it’s worth. Oh, and by the way, do you have a pilot’s license? You’ll probably need one. And you’ll have to learn where all the local no-fly zones are. Can’t have you flying into the path of an Airbus.
Third, considering the way most people drive and the way most people maintain their automobiles, do you really want them piloting their sky-cars overhead? Considering the way some people use their vehicles for dangerous street-racing and the way other people keep getting themselves into road-rage duels, do you really want that going on overhead? Considering the way some people throw trash out the windows, do you want them overhead ? And finally, do we really need another source of noise and pollution in the air?
And all of the above assumes that the engineering problems can be solved. The Moller sky-car has been in development for how long? Since 1974. And it’s still “just a few years away.” (Critics say that they have yet to solve significant noise and stability problems.)
Flying is not the same as driving and any good pilot will tell you that it’s a whole other mindset— not just a set of skills, but a discipline . Including a discipline of maintenance . Considering how most people do discipline…uh, no, I just don’t see this one happening any time soon. Okay, maybe eventually on a small scale, maybe as flying taxis from local hub to local hub, maybe—but as a mass-production item? Not likely. I think that most people will probably invest in more cost-effective travel.
Imagine a printer that operates in three dimensions, building up solid objects a layer at a time. Such printers exist and are used for making prototypes and models. Depending on what kind of material can be layered and the resolution of the printer, it might be possible to print up objects as mundane as toasters or as rare as star sapphires. (We might not have to wait for Robby the Robot to crystallize the gems.)
But even more important, we’re on the threshold of being able to fabricate living tissue. Researchers have already demonstrated that they can print living cells onto a collagen framework to create specific tissues and even whole functioning organs. We might eventually be able to grow our own replacement organs in the lab—skin, hearts, lungs, kidneys, livers, ears, hands, feet, arms, legs—and not have to wait for some unfortunate motorcyclist to lose an encounter with an SUV. We could see this happening within ten years. Could we grow whole new bodies…? We won’t know until we get there, but once upon a time a heart transplant was unthinkable too.
Beyond that, being able to print living tissue could revolutionize agriculture. Why breed a whole cow when you can grow a steak in a bio-fab factory? Once the process is perfected and the product is approved safe for human consumption, a bio-engineered filet could be cheaper, safer, and healthier than meat produced the old-fashioned way. And a lot more humane. But why stop at steak? We could grow any cut of meat we wanted, and probably far more economically than raising a whole animal. Want some fresh dolphin or whale meat? Elephant? Panda? (Even cannibals might be able to legally … never mind.)
Of course, we’d still maintain herds of all kinds for genetic diversity, but we wouldn’t need to destroy the rain forests of the world to create more pasture for more cattle to feed the world’s growing appetite for meat. This one is a no-brainer. It’s not just a growth industry, it’s a growth industry. As the world’s population continues to grow, factory farms may be our only hope for avoiding a food crisis. We might see this before 2020.
Internet Protocol Version 6 is already here. We’re switching over now . Prior to IPV6, internet addresses were limited to 32 bits. Under IPV6, internet addresses are 128 bits. This means that there are now 2128 possible internet addresses (340 undecillion ), or in more understandable terms “umpty hella-gazillion”—enough so that every living human being on the planet could have 5*1028 separate and specific domains.
What this means in practice is that every thing on the planet worth anything at all, manufactured, grown, discovered, studied, observed, or born , can have its own web address and associated locater-chip. Can’t find your car keys? Just ask your phone where they are. Want to know where your steak came from, what lab it was grown in, what nutrients were in the tank, and who inspected it? That’s available too, ask your phone.
Your car will be able to drive itself so you can talk on the phone, read a book, or watch TV—it will converse with the vehicles around it, informing them when it needs to change lanes, and all the cars will adjust to maintain safe distances. Want to know where your teenager is at 12:30am? You’ll be able to track his location easily—and if he’s out street-racing, you’ll have evidence of that too.
Want to know how much cash is in your wallet? Ask your phone. Why is there a twenty missing? Your phone will tell you that one of the twenties was removed from your wallet while you were in the shower and is currently in the pocket of your sixteen year-old son. Want him to come home now? Tell the car to bring him home safely.
Had your purse stolen? Ask your phone to alert the police. The thief will be picked up momentarily. Had your car stolen and taken to a chop shop? The police will know where every single piece of it went.
Just bought insurance and need to inventory your physical property for a rate adjustment? Ask your phone. You can print out a list of everything you own, when you bought it, how much you paid, what it’s worth now, and what the replacement cost would be in case of fire, flood, earthquake, tornado, or asteroid impact.
Can’t find your phone? Ask the refrigerator.
But wait, it gets better. Humans will be chipped too, just like dogs, cats, and cattle. Can’t remember the name of that little restaurant you liked in New York? No problem, your personal life history is stored in the cloud. We can remember it for you wholesale. Sign up for Apple’s iMemory service.
Catching rapists, muggers, thieves, and murderers will be a lot easier. The cloud will maintain a location-tracking service of everyone, chipped or not. There will be cameras everywhere. Court trials will have a whole new level of evidentiary standards.
Here’s how The Singularity will happen.
There’s this thing called “emergent behavior.” It means that complex patterns and events can arise out of relatively simple interactions. One ant is one ant, but a whole colony of ants behaves like a gigantic multi-cellular organism—that’s emergent behavior. One car slows down at a curve in the highway at five in the morning, it’s one car—but twelve hours later, when there are hundreds of cars on the same highway, you get a standing wave in the traffic flow, a wave that actually travels backward from the source—that’s emergent behavior. One person goes to the bathroom and flushes the toilet, no problem—but in the early days of television, when I Love Lucy broke for a commercial and a million New Yorkers all went to the bathroom all at the same time, the reservoir levels visibly lowered—that’s emergent behavior.
When the whole world is linked in a massive network of chairs and trash cans and lawn mowers and refrigerators and cars and streetlights and smartphones and supermarket packages and pants and skirts and underwear and even shoes and socks (no more lost socks?!)—when the whole world is totally immersed in a global web of interconnections, when even dollar bills are monitored for their travels through the economy, there will be emergent behavior .
All of these separate chips will have software-specific functions. Your shirts will tell the wash machine how they want to be washed. Your frozen dinner will tell the microwave how long it needs to be cooked. Your shoes will tell you when they need to be re-soled. Your internal monitors will warn you of diabetes and gout and heart disease. The menu at the restaurant will advise you on your healthiest choices.
All of this data—not just yours, everyone’s —will get sucked into the cloud, massaged, shared, digested, fiddled and diddled, jiggled and juggled, sorted and ported, creating the most accurate real-time census possible. Trends of all kinds—social, political, economic, cultural, biological—will be recognized first by the cloud and responded to even before humans are aware. The buying habits of millions will tell industry just how many boxes of Cheerios to produce and how many Kinects to manufacture. A super-intelligent cloud will advise producers whether or not it’s cost-effective to produce. Before you go shopping for a car or a house or even a new TV, your phone will let you know if you can really afford it. And even more personal, your health-care will be automatically triaged based on the availability of doctors and based on your previous record of cooperation with preventive medicine.
The super-intelligent cloud won’t have consciousness as we understand it. But if Marvin Minsky’s theory ( Society Of Mind , by Marvin Minsky) is correct—that sentience occurs as a product of multiple interrelated subroutines—then eventually the super-intelligent cloud will begin to function not just as a monitor of all the data flowing through it, but as a mentor as well.
The Singularity—as I see it—will inherit all the best and all the worst traits of the species that produces it. It will come into existence as an assemblage of software mechanisms designed to serve our fundamental wants and needs, but it will evolve . And because the essential goal of life is to survive, it will most likely evolve into a symbiotic consciousness with humanity. And if that happens, then it will have a built-in bias to keep us functioning at our best.
We’ll see. It will happen in our lifetimes.