An article covering the monumental story of an AI beating the world champion Go player discusses the societal ramifications: Deep Learning Is Going to Teach Us All the Lesson of Our Lives: Jobs Are for Machines. The program Alpha Go beat the human Lee Se Dol. Some of the ramifications:
The article presents jobs across two dimensions: routine and non-routine, cognitive and manual. Routine jobs in both cognitive and manual have stopped growing starting around 1990. The assumption/premise, and not a controversial one, is that this is a result of automation. The question, and this is the crucial, has become when will non-routine jobs succumb. As a software engineer, I tend to think that software engineering will survive (as did Go players...) as well as health care, based basically because I think non-routine cognitive jobs will survive the longest. And I still have the bias that human-to-human interaction will retain its value in some industries.
AIs referenced in the article: Amelia call center AI. She replaces call center workers across multiple languages and learns faster than any low-wage human. The Viv personal assistant which promises to bypass the ubiquitous advertising that pays for web content. These subjects are regularly discussed in the subreddits Futurology and DarkFuturology. So the subject of post-scarcity mediation is not something for the distant future. As jobs progressively become not-done by humans, the question of how do humans get money to pay for materials (produced by machines?) becomes more common. When there are no jobs for humans, how do they pay for food etc.? The basic income proponents are already thinking about that and Switzerland is already preparing.
Is the blacksmith analogy dead? The confidence that, just like smithees had to adapt to the elimination of their jobs 150 years ago, current jobs will morph into a new job market. Maybe there are no new jobs to replace the eliminated. And, as a developer, my whole career has been eliminating jobs.
Valuable reading is 20 Crucial Terms Every 21st Century Futurist Should Know. It suggests a radical socialist idea to me: The move from some automation to complete automation starts with a redistribution of wealth to those in control of the automation. It initially appears as the class struggle. What happens to those wealthy and poor when the economy collapses?
Listening to dialog from the movie The End of the Tour (which we will probably go see tonight), spoken by David Foster Wallace:
So look, as the internet grows in the next 10, 15 years, and virtual reality pornography becomes a reality ... we're gonna have to develop some real machinery inside our guts to turn off pure, unalloyed pleasure or I don't know about you, I'm gonna have to leave the planet. ... 'Cause the technology is just gonna get better and better and it's gonna get easier and easier and more and more convenient and more and more pleasurable to sit alone with images on a screen given to us by people who do not love us but want our money and that's fine in low doses but if it's the basic main staple of your diet you're gonna die.
Taken from the On The Media interview that Brooke Gladstone had with David Lipsky 'We've Sort of Become Friends': The Original Tapes from David Foster Wallace's '96 Book Tour (transcript here).
This fear is examined and realized in Peter Watts' book Echopraxia in several forms: biological sex has become non-existent, relationships are created by altering each others' personalities to be more compatible, a large percentage of the human population lives only hooked in to a virtual world, etc. He suggests that our basic desire to eliminate pain will inevitably lead us to become not-human. Our toolmaking will destroy us. DFW worried in a similar manner concerning drug addiction and pleasure-seeking. The idea of a rat continually pressing a button for pleasure until it dies is unnerving. When I had written about Echopraxia, I recognized the connection to works by several other SF writers but hadn't considered how deeply these themes are encoded in Infinite Jest until I heard the quote above.
I'd recently read Blindsight by Peter Watts after my Kindle recommended it and the content of the reviews suggested I would like it (correct and correct). It's a posthuman sci-fi novel about a small group of enhanced humans sent on a decades-long journey to investigate an alien presence at the outer regions of the solar system. Throughout, there's a sense that humanity has reached a dead end. As people with means alter themselves with enhanced abilities--man/machine interfaces, multiple consciousnesses in one person, neuro-physical updates--un-altered "baselines", without the ability to keep up against an advanced society, hook their brains up to a virtual world called Heaven. Once in, they abandon any emotional bond to those they left behind. One particularly odd aspect of the novel is that vampires exist as an extinct offshoot of humans. They have been resurrected and though there is a detente of sorts, they are so far advanced in intelligence and ability that even the most enhanced humans are like children. To trump even this level of insignificance, the aliens the crew encounter are orders of magnitude more adept than even vampires.
I've since started the sequel titled Echopraxia (which means "the involuntary repetition or imitation of another person's actions"). I'm 50-or-so pages in, but the grimness is the same. The action takes place on Earth where a group of hive mind Bicamerals push the boundaries of invention but cannot explain how they achieve it. It's another example of humans becoming so far from equal that there is no longer a single humanity. Non-Bicamerals are as threatened by them as baselines by the transhumans. Similar fears have been echoed in Bacigalupi's The Windup Girl (and his even more stunning and depressing short stories Pump Six), Rajaniemi's The Quantum Thief trilogy, and even somewhat in Atwood's Oryx and Crake trilogy. Notably, Watts is a marine biologist, Rajaniemi a mathematician and programmable DNA entrepreneur, and Atwood a developer of remote robotics.
Contrast these with Vernor Vinge's Rainbows End (read at the same time as Oryx and Crake). Though he deals only slightly with posthumanism, he shows the tension of modern, augmenting technology. Rainbows End has all of the aspects of a dystopian warning--aggressive emergent AI, near 100% surveillance state monitoring, the physical destruction of all books in order to digitize--yet he somehow offers an optimistic message in the end. Ever the singularitanist. Or maybe the realist. Still, sci-fi is not about what will happen; it is simply plotting a straight line with a few of the data points we currently have. Sometimes it's a warning, sometimes a hope.
Several months ago, I finished the third book in Hannu Rajaniemi's post-human Jean le Flambeur series. It's a monument of fiction of what-will-happen-when-we-can-upload-consciousness that has social cachet for today's various technological anxieties. Fantastical art as commentary on present day. While researching author interviews and secondary sources after-the-fact, I read one of the author's recommended source articles titled "If Uploads Come First", written in 1994 by Robin Hanson. One key dramatic history, revealed near the end of the Rajaniemi trilogy, was that when brain uploads happened humans became dupe-able VM workers for tasks that algorithms couldn't accomplish. What a computer can't achieve, 5K uploaded brains could be sold into slavery to achieve. The problem that computers can't do X is solved by simply having digitized humans do it by the thousands.
A few years ago, a coworker told me about out-sourced personal assistants. These are 3rd-worlders who will sell their time by the penny to organize your life remotely as best as possible. The concept was as fascinating as it was morally ambiguous, but I never took advantage. How to organize such unlimited resources? This I had encountered before--conceptually--with Amazon's Mechanical Turk. Again, not having partook, the MT offers a marketplace of people-who-will-do-anything with people-who-need-anything-done. Think 1000s of addresses to enter for fractions of pennies an address. People have time and need money, and there is work to do that would be too costly to write a computer program for.
Friday was an argument with the boss about whether a large VM containing multiple server instances was more valuable than multiple, smaller VMs of equal summed potential.
I'd read Rajaniemi's first book in this series, The Quantum Thief, a little over a year ago and have been anxious for the follow-up. Where the first riffed on themes of presence as identity--common in posthuman lit--this worked through story as identity, modeling it's chapters after those stories in The Arabian Nights. This has the same fire hose of information as the first but with a different character: the primary setting is Earth gone desert and infested by nanotechnology that can both steal your mind (through stories) and subvert your body to destructive growth. Perhaps more moody than the first?
It's an effort not to be irritated by a book that demands multiple readings and earns them.Continue reading "The Fractal Prince; Hannu Rajaniemi"
Highly recommended heist/detective novel dressed in post-human scifi garb. Our hero gets busted out of a dilemma prison by a warrior from the Oort cloud in order to retrieve a valuable object that a previous version of him had hidden on one of the walking cities on Mars where people use life time as currency before they are harvested by the Resurrection Men to become Quiets, human/machine hybrids that sustain the city and terraform the planet. Not for the technologically squeamish. This was a completely show-don't-tell novel that, despite the maelstrom of undefined terms, provided thrilling action next to thoughtful drama. Another great recommendation from the io9 book club. Moving on to The City and The City next.
See the Wikipedia entry for more details plus keep their articles on characters and terms used for a reference while reading. [ updated 22 Mar 2015 ] The terms page has been deleted, but has thoughtfully been archived by Karan Gill here.Continue reading "The Quantum Thief; Hannu Rajaniemi"
Read these to take a break in between the first and the second Stieg Larsson books. With The Chaos Scenario, one of your two two favorite hosts of On The Media riffs on the various disruptive aspects the internet will have on media, business, and, ultimately, society as a whole. There were many questions and few answers but rather cautionary tales. If you like his discursive style on OTM, you'll enjoy his very conversational writing style.
Kurzweil's book presents a conundrum: how to objectively approach a 10-year-old book that attempts to predict the world 10 years into the future (and much further)? Additional: how to give fair judgment when the author emphasizes their credentials yet over-estimates potential accomplishments within those credentialed areas? This was an impulse purchase after reading it referenced with praise in several discussion threads. Although many ideas were interesting, I came away disliking his inelegant, artless writing style and generally dry structure. In contrast, I thoroughly enjoyed the predictions and explications of bioengineering in Fukuyama's Our Posthuman Future, even though I disagreed with his conclusions.
[ update 30 Nov 2010 ]
Ray Kurzweil's Slippery Futurism from IEEE Spectrum [ via Slashdot ] gives a drubbing to predictions in The Age of Spiritual Machines, examining both the difficulties in separating what had been common knowledge 10 years ago from what seems prescient now, and the difficulties in getting Kurzweil to admit when he was, obviously, wrong. The inexplicable crowing that Kurzweil does in the book regarding his past business successes seems more explicable now.
Heard about this via the io9 book club and so purchased the hardback of it and his short story collection Pump Six. TWG has been praised by Time Magazine as one of the top ten books of fiction for 2009 and by the American Library Association as the best SciFi of 2009. Halfway into the story thus far and it feels very of-a-time with Naomi Klein's and Michael Pollan's ideas, along with (in a more minor fashion) Fukuyama's somewhat older book Our Posthuman Future.
Listened to Toby Miller, author of Makeover Nation: The United States of Reinvention on Media Matters (stream available). He discusses, among other subjects, the role of direct-to-consumer marketing of drugs and their use to alter personalities as one of the methods that America has, dubiously, excelled at using to reinventing itself. Similar discussions are echoed in Fukuyama's book Our Posthuman Future (see also my reference in in July 2005). Miller worries that people are presented with too much power with the direct marketing. He argues that just as the unskilled should not have the power to build bridges or airplanes, they should not be presented with the power to choose their medications. Not the perfect comparison but it gets to the point. Similarly, Fukuyama warns that as we have more freedom to alter our personality (he complements this with a discussion on our power to alter our physical self), we take away apparent deficits that are actually benefits. One of the driving forces behind human accomplishments is those "negative" feelings of fear, unhappiness, inadequacy. Take those away and we have less drive to conquer the objects of our distress.
Maybe we're still in a nascent age of mood-altering drugs. In the same way that there is an older generation who cannot adapt to using cell phones, text messages, and mobile internet--a technical alteration that gives us a connective and communicative power--there will be a generation that can't adopt the use of mood alterations that may give us concentrative and expressive power. Such drugs certainly don't exist today, but their potential to exist in the next 20 to 30 years is feasible. Once this is accessible and non-addictive, won't adoption of its use be as common as adoption of current technology? Once common to some, the gulf in ability between those enhanced and not could be similar to that of those born with rare physical prowess and those not. We could see an enhanced class of people who would be working on a much higher plane of consciousness with the rest of the population--most likely poor--fumbling around unfocused.
Scientists have been injecting human brain cells into monkey fetuses (NEWS.com.au, LiveScience.com, Google News). The resulting animals, if left to grow, could develop self-awareness--or, if as some argue, many animals are already self-aware, they could develop consciouis characteristics more indiscernible from humans.
Last November, I noted that animals were begin engineered to produce more human-like organs. Fukuyama's Our Posthuman Future [Amazon] addresses many of these issues and suggests that limits be defined soon so that the "natural" definition of human is not altered.
Earlier this year, Marshall Brain (yes, Brain) of the wonderful How Stuff Works put out a longish essay titled "Robotic Nation." In it, he predicted 50% unemployment in 20 years as automation takes over the workplace. As it stands, the world's economy would collapse.
He's been busy since then bolstering his argument:Continue reading "Robotic nation"