Robot Apocalypse, Robot Utopia

There's no doubt in my mind that we're heading into what's called "the knee" of an exponential curve in terms of robotics  - that moment when the accumulation of molecular advances in the technology and everyday penetration of a piece of technology takes off. As I discussed in a previous post the evidence for this exists in a number of places, with the most recent affirmations (to my mind) being the announcement of a general purpose personal robot - the Luna - to be launched at the end of this year and the near-term intention to launch an industrial, multipurpose robot at a price point around $5,000, making it affordable for small companies. But there are many other signs that this is the case - advances in autonomous swarm robotics to allow teams of small robots to explore disaster sites; robots trained to develop their own language to describe their experience to other robots; advancing robot vision systems using biological modeling of a visual cortex in a monkey. And that is just a small sample of the many convergent technologies that are acting as enablers for the coming robot revolution.

What will the robot revolution mean for human society? How will we be changed by the arrival of ubiquitous robotics? What will it mean in terms of everyday life, in the home, in the workplace, in healthcare? Generally, of course, we are given two models of possible futures in a robotic world - the apocalypse a la Terminator and the Matrix and the "utopia" with helper robots a la Astroboy or Data on Star Trek TNG. In one instance robots, as proxies for all technological advancement, finish us off as a species, in the other technology is seen to solve all of our problems - world hunger, poverty, climate change, etc. However, I want to suggest that the question is rather more complicated than that.
First, a little mea culpa: I love technology. My wife and I each have a smartphone and Macbook Pros and an Apple TV. My wife - a photographer - shoots with DSLRs, one of which also shoots HD video. I regularly read the tech press - from the techno-cyber-philia of the transhumanist websites to the more "sober" pages of MIT Technology Review, New Scientist and IEEE Spectrum. We desperately need to renovate our crumbling kitchen and part of me hopes we don't get the money until its possible to get an internet connected smart home that I can access from my phone anywhere in the world and that has augmented reality projectors in the kitchen. You get the point - I love tech.

But loving the stuff doesn't mean accepting the promises and threats that are thrown about uncritically. It is simply not true, for instance, that the problems of poverty and world hunger remain unsolved because we have to achieve the level of technological advancement required to feed and house people. There is more than enough food produced (I won't get into the quality of the food produced - high carb cereal grains providing the bulk of global crops, for instance) but that food for which there isn't a market - i.e. people with money to purchase it - is allowed to rot or is dumped at sea. Many countries in which there are the worst problems with food shortages are actually net exporters of cash crops - whether bananas, coffee beans or cacao.

Even if there weren't a surplus of food crops it is also not the case that there is a lack of resources to convert non-arable land, provide seed stocks, fertilizer, farm machinery, etc. At present the USA alone spends close to $1 trillion on various aspects of the military and intelligence apparatus. The next ten military purchasers combined spend another $1 trillion. That is more than enough to feed every hungry human, house every shelterless family, provide clean water and teach every illiterate to read. Lastly, the technological priorities that do exist - which make the private automobile the pre-eminent mode of transportation, encourage short distance flights for business, expand carbon-heavy low density development (i.e. suburbs), make use of fossil fuels for our power plants - contribute to climate change, which exacerbates food shortage in many place. The so-called solutions of converting grain stocks and arable land to the production of biofuel only contributes to the problem.

The problems of scarcity are thus social problems and not technical ones. And having a robot in the home or in the field or the factory will not solve those problems. The irony could be that agricultural workers could be displaced by robots and then not have the income to buy the food that the robots are able to produce in greater quantities. And here again, the problem isn't that robots are taking jobs - it is that the priorities of our society are such that rather than being used to make our lives better - as tools (whether hammers or robots or computers) are meant to do, they instead threaten to make our lives worse at least in the short term. But don't blame the robots - blame the Man.

Next part in this series: Will Robots Lead To Revolution?

The Robot Invasion Has Begun

Lock up your family, protect your job! You don't have to be Ray Kurzweil to see that there has been an acceleration of certain kinds of technological progress in the last generation - the rise of the PC, followed by cellphones, the internet, laptops, smart phones (only since 2007!) and now tablet computers. There have also been advances in machine vision and materials technologies - witness the big leap of the Xbox Kinect, which, along with touchscreens in smartphones and tablets, is radically reshaping our concept of human/machine user interfaces. All of this seems to have laid the groundwork for a near-term revolution in robotics. And such a revolution is in many ways overdue. Robotics pretty much stalled after the introduction of industrial robots into the large manufacturers, such as the auto industry, in the 60s and 70s. However, in the last number of years it has slowly been making advances that have created the conditions for a new leap forward.

Already, there are a number of popular open-source research platforms, like the PR2 from Willow Garage (something of a spin-off from Google) that is in the video above demonstrating a new software module that enables the PR2 to pick up and scan the bar codes of groceries (bye-bye check-out clerks?). The makers of the very popular Nao robot from France are about to go open-source to speed up the process of developing software "apps" to supply behaviour modules for the cute, versatile little robot. Google has also announced that it will extend its Android open-source software's capabilities into the field of robotics with Android@home, which will provide a common operating system for the growing interest in home management or smarthome systems - things like dimming lights, controlling the thermostat, home security, cleaning robots, etc. And there are popular events like the Robocup, that brings together programmers and robot aficionados to engage in a robot football tournament with the goal of creating better-than-human robot football/soccer players by 2050. To be honest, I think it will be sooner than that.

All of this stuff is interesting and even exciting, however it has mostly been about laying the groundwork and robotics is waiting for the killer device to make the breakthrough into ubiquitous robots in the same way that Apple made the breakthrough for smartphones and tablets. There are two immediate potential contenders in this field - Willow Garage is perhaps a contender but at present their interest is still in the field of research and development platforms with a $400,000 robot that's out of reach to consumers and most businesses. One is Heartland Robotics, headed up by Rodney Brooks a pre-eminent robotics professor, formerly of MIT, and founder of iRobot, which brought the Roomba vacuum cleaner to market. Of the 8.6 million robots in the world, something like 7 million of them are Roombas, so he has a track record of knowing how to convert existing tech into a consumer device. Brooks now intends to take that success into the field of light manufacturing by advancing on the platform developed with the Obrero (worker) robot - basically an intelligent, dextrous arm and hand - from a few years ago. While Heartland is in what's called stealth mode - primary research to complete development of the product out of sight of competitors - the general goal of Heartland is public knowledge.

Brooks wants to create a robot for light manufacturing that will be priced in the $5,000 range, allowing the USA to compete with low wage economies to which a lot of manufacturing has been outsourced - places like China and Asia. I haven't done the math on the relative advantage of a $5,000 manufacturing robot vs a product produced by sweated labour plus the costs of shipping and handling but my guess is that it would make a lot of domestic manufacturing more competitive. Brooks believes that this will generate a new industrial revolution and will eliminate drudge work in the service sector, particularly in health care. The first impact, however, is likely to be unemployment in those industries that can make use of a cheap robot. There will be other impacts, such as the long term decline in the rate of profit once affected industries take up the new technology across the board and the effect of relative advantages is mitigated (I'll take up some of this in a future post). Whatever the ultimate character of the impact, it would no doubt be huge. Check out 32 minute Brooks talk, below, from the Maker Faire in 2009:

The second potential "killer device" is the forthcoming general purpose home robot, Luna. While we have the Roomba in the home, there has yet to be any general purpose robots to perform a variety of tasks. I'm not sure that the Luna is it - it seems to have too little functionality. But the intended future price tag of $1,000 (for this year and next it will be $3,000) might be low enough to encourage enough early adopters with disposable income and a desire to be the first ones on the block to own a home robot to make it a viable business model. The Luna platform is intended to be expandable both in terms of hardware and software, with the idea that developers - in the model of the iphone and Android app stores - can create specific modules to expand the capabilities of the Luna - say voice recognition software or a more advanced arm, etc.

But even if the Luna itself isn't ready for prime time and doesn't take off as the company hopes, it is a sign that the genie is out of the bottle - the age of personal robotics has begun. To be honest I haven't grokked what the cultural implications of a robotics explosion will entail - it certainly won't mean the end of inequality, world hunger or imperialism - most of the "service" robots other than the Roomba are for military use to defuse IEDs, as spotters, UAVs and, already in testing, the next generation of fully robotic fighter planes known as the Phantom Ray. But it would be foolish to think "nothing will change" and the coming robot revolution deserves some serious social analysis. Thoughts?