Saturday, June 21, 2014

Einstein may have discovered dark energy without even realizing it

By George Dvorsky
An historian of science from New York University has re-interpreted a correspondence between Albert Einstein and Erwin Schrodinger in which the two scientists argued about the nature of the cosmological constant — a kluge that Einstein embedded in his general theory of relativity to explain why the Universe was neither expanding or contracting (what scientists thought at the time). According to Alex Harvey, without the two of them even realizing it, they were actually discussing the potential for dark energy — an idea that wouldn't hit the cosmological radar for another 70 years.
Back when he proposed the general theory of relativity in 1916, Einstein didn't know that the universe was expanding. In his mind, the size of the cosmos was fixed — what was being held in place by this thing he called the 'cosmological constant.' Without it, the universe would have contracted or expanded in accordance to the amount of mass within it. This mysterious force, argued Einstein, was what held everything together in place.
Of course, he had to revise his idea during the late 1920s after scientists discovered that the universe was in fact expanding. After he removed the constant from his equations, Einstein referred to it as the biggest blunder of his career.
But as history now knows, it wasn't that bad of a blunder. In fact, he was actually on to something.
As Alex Harvey has now revealed in a recent paper, Einstein, in conversation with Schrodinger, described a characteristic of the cosmological constant that is now regarded as a fixture of dark energy theory — the idea that it's a non-uniform force in the cosmos that's pushing everything outwards. Moreover, he dismissed the idea right there and then on account of the problems it would create for physicists trying to quantify it — something that has most certainly happened.
Specifically, Einstein and Schrodinger were discussing the properties of the cosmological constant and speculating about what form it might take. Schrodinger wondered if the "cosmic gravitational field" would be fixed or varied in terms of its strength, to which Einstein responded:
This means, one not only has to start out from the hypothesis of the existence of a nonobservable negative density in interstellar spaces but also has to postulate a hypothetical law about the space-time distribution of this mass density. The course taken by Herr Schrodinger does not appear possible to me because it leads too deeply into the thicket of hypotheses.
Bingo. As Harvey notes in his paper, "Einstein described not only the central problem of the search for dark energy but the headaches in formulating its structure." Indeed, Einstein predicted the exact problem now confronting cosmologists as they struggle to devise a coherent theory that explains the exact mechanics of dark energy. They know it's there — but exactly how it works and how it's proportioned throughout the cosmos is a complete mystery.
You can read the entire paper here.
Top image: Dark energy graphic via NASA/STSci/Ann Feild
Reference: io9

Monday, June 9, 2014

Get ready, robots are going to steal your job

Federico Pistono, founder and CEO of Esplori
Are robots stealing our jobs? Is human labor destined to become obsolete? This is a scary topic that has been debated for more than 100 years. Economists even created a term for this—the Luddite Fallacy—referring to the 19th-century English textile artisans who protested against newly developed labor-saving machinery. But fast-forward to today and you can see a startling trend emerging. Almost unnoticed, computer power is growing exponentially, and it is advancing efforts to mechanize the labor force.
Conventional economic theory suggests that for every job displaced by technology, new jobs and new sectors are created. Historically, this has been true. When we moved out of the farms, we started working in factories, and when human-powered factories became mechanized, we invented the service sector. We adjusted. But today "technological unemployment," the term used to describe machines, robots and algorithms replacing human labor for good, is starting to feel less like a far-fetched idea and more like reality, and it has its roots in the exponential nature of technology.
Computer speed (per unit cost) doubled every three years between 1910 and 1950, doubled every two years between 1950 and 1966 and is now doubling every year. What used to cost hundreds of millions of dollars and took up entire building now sits in the palm of your hand and costs a hundred dollars or less, and it all happened in a few decades. In fact, chances are that you are using such technological marvel to read this very article.
This exponential nature of the growth of technology has ramifications far beyond mere computer speed. Entire fields are beings revolutionized. Here are some examples:
The first sequenced human genome was complete in 2003 at a cost of nearly $3 billion, and it took 13 years. Just a decade later we can do the same in a few days for less than $1,000.
Industries in Transformation
The artifically intelligent computer system known as IBM Watson is now entering the health sector by leveraging its natural language—hypothesis generation—and evidence-based learning capabilities allow it to function as a clinical-decision support system for use by medical professionals.
Companies worldwide are planning on going fully automated by using advanced robots in their production line. Foxconn Technology—the world's largest electronics manufacturer, based in Taiwan—has already installed hundreds of thousands of robots to replace human workers, with a goal of moving toward 1 million. Japan's Canon is also ditching human production-line employees, so it will rely entirely on robots to build its cameras by 2015. And algorithms are already writing articles on real estate, financial analysis and sporting events at a quality that is indistinguishable from that of a typical human journalist.
The first fully autonomous vehicle was announced by Google on March 1, 2012, when the Nevada Department of Motor Vehicles issued the first license for an autonomous car. Experts in the sector claimed a decade would pass for another company to catch up. But just two years later all major car manufacturers have announced working prototypes of fully autonomous vehicles and already are planning commercialization.
What I have described are market shifts going unnoticed by most as industries morph at a faster and faster pace. In the not-too-distant future, we might be looking at a fully automated production and distribution line in almost any sector.
Consider food. A network of autonomous tractors, swarm robots and sensors can grow, collect, separate and package food while performing quality control. Drones will constantly monitor the fields from the sky and give instructions to the robots on the ground. Autonomous vehicles will ship the food wherever needed.
Robots and algorithms will manage distribution in warehouses, cook food and sell it through automated kiosks. All the individual pieces of this technology already exist today, either at the prototype or at the commercial level. The same concept can apply to virtually any industry, given enough time.
Last September, Oxford University released a study estimating as much as 47 percent of U.S. jobs are at risk of being replaced by technology within the next 20 years. Many economists are supporting this thesis as a concrete possibility, since we're now at a point where there is too little time for us to adjust to these paradigm shifts.
There is concrete evidence. The U.S. Bureau of Labor Statistics released this data that shows the civilian labor-force participation over time against corporate profits. Gray areas indicate recessions. The green angles represent the strength of the recovery in relation to the number of people employed in the economy.
As we can see, corporate profits have not been affected; in fact, they are at an all time high, while job recovery has been shallower and shallower each time. The employment-population ratio is at its lowest since 1983. Before that, women had not entered the workforce in large numbers, so comparisons to today are meaningless.
This is called a jobless recovery, and there are multiple reasons for it. Outsourcing certainly has played a role, but as Foxconn shows, we are already seeing a reverse effect. The main culprit is a sharp increase in productivity, part of it being better processes and managerial decisions, but mainly technological progress. Automation, computing power, robotics, machine learning, data analysis, smart algorithms—you name it.
However, I believe that technological unemployment—which represents a structural and irreversible trend in unemployment, as opposed to a cyclical one—is not an inevitability. I'm sure that potentially we can come up with millions of new but unnecessary jobs in the future. These are jobs that drive GDP growth vs. creating value for society. Just a glance at what we have accomplished in the last 50 years should be enough to make that argument very credible, indeed.
But have we ever considered the possibility that finding replacement jobs, no matter what they might be, could be the wrong choice to begin with?
Looking at the data, I see the convergence of two worrisome trends: dramatic increase in inequality and technological innovation displacing workers like never before. Given these two conditions, I think we need a serious discussion at a global level about what the purpose of the economy should be. I think we need to renegotiate the social contract between the people, the state and the private sector. This discussion needs to happen now, before things escalate and before millions of middle-aged, unskilled workers find themselves out of a job, with no hope of getting another.
Governments, companies and people should prepare for this paradigm shift with the realization that there is no silver bullet that will magically solve everything and that things don't work in silos. The policies that a government decides to enact will affect its citizens and the private sector. These can include tax reforms to provide a safety net for those who become unemployed, smaller government, programs to stimulate start-up innovation, sharing and open source and providing citizens with an unconditional basic income (a federal stipend guarantee). I hope more people join the debate to help find the right solutions.
Federico Pistono, founder and CEO of the online learning start-up Esplori, is a computer scientist, activist and social entrepreneur. He is also the author of the book "Robots Will Steal Your Job, But That's OK: How to Survive the Economic Collapse and Be Happy" and a Singularity University graduate.
Reference: CNBC