Sense and Nonsense Of interest to???????????? Mostly stolen from other truly clever people.
Saturday, June 4, 2016
SINGULARITY IS NEAR
Ray Kurzweil speaks on ‘Singularity’ in San Francisco, 07 February 2007. AFP PHOTO/GABRIEL BOUYS (Photo credit should read GABRIEL BOUYS/AFP/Getty Images)
When Ray Kurzweil published The Singularity Is Near in 2006, many scoffed at his outlandish predictions. A year before Apple launched its iPhone, Kurzweil imagined a world in which humans and computers essentially fuse, unlocking capabilities we normally see in science fiction movies.
His argument though, was amazingly simple. He pointed out that as technology accelerates at an exponential rate, progress would eventually become virtually instantaneous—a singularity. Further, he predicted that as computers advanced, they would merge with other technologies, namely genomics, nanotechnology and robotics.
Today, Kurzweil’s ideas don’t seem quite so outlandish. Google’s DeepMind recently beat legendary Go world champion Lee Sedol. IBM’s Watson is expanding horizons in medicine, financial planning and even cooking. Self driving cars are expected to be on the road by 2020. Just as Kurzweil predicted, technology seems to be accelerating faster than ever before.
Reason #1: We’re Going Beyond Moore’s Law
For the last 50 years, the technology industry has run on Moore’s Law, the famous prediction made by Intel co-founder Gordon Moore that the number of transistors on a microchip would double about every 18 months. That’s what enabled computers the size of refrigerators to shrink down to devices we can hold in the palm of our hand.
Now we are approaching the theoretical limit and the process is slowing down. The problem is that you can only shrink transistors down so far before quantum effects between atoms cause the transistors to malfunction. While chip technology is still advancing, at some point you can’t cheat mother nature anymore. Moore’s law will come to a halt sometime around 2020.
Yet Kurzweil has pointed out that microprocessors are in fact thefifth paradigm of information processing, replacing earlier technologies such as electromechanical relays, vacuum tubes and transistors. He also argues that the numbers of transistors on a chip is a fairly arbitrary way to measure performance and suggests to look the number of calculations per $1000 instead.
The first industrial robot, Unimate, first arrived on the GM assembly line in 1962, welding auto bodies together. Since then, automation has quietly slipped into our lives. From automatic teller machines in the 1970’s to the autonomous Roomba vacuum cleaner in 2002, machines are increasingly doing the work of humans.
Today, we’re beginning to reach a tipping point. Rethink Robotics makes robots likeBaxter and Sawyer, which can work safely around humans and can learn new tasks in minutes. Military robots are becoming commonplace on the battlefield and soldiers are developing emotional bonds with them, even going as far as to hold funerals for their fallen android brethren.
The future will be more automated still. The Department of Defense is already experimenting with chips embedded in soldiers brains and Elon Musk says he’s thinking about commercializing similar technology. As the power of technology continues to grow exponentially—computers will be more than a thousand times more powerful in 20 years—robots will take on even more tasks.
Reason #3: We’re Editing Genes
In 2003, scientists created a full map of the human genome. For the first time, we actually knew which genes were which and could begin to track their function. Just two years later, in 2005, the US government started compiling the Cancer Genome Atlas, which allows doctors to target cancers based on their genetic makeup rather than the organ in which they originate.
Now, scientists have a new tool at their disposal, called CRISPR, which allows them to actually edit genes, easily and cheaply. It is already opening up avenues to render viruses inactive, regulate cell activity, create disease resistant crops and even engineer yeast to produce ethanol that can fuel our cars.
The technology is also creating no small amount of controversy. When you start editing the code of life, where do you stop? Are we soon going to create designer babies, with predetermined eye color, intelligence and physical traits? Should we alter the genome of mosquitoes in Africa so that they no longer carry the malaria virus?
These types of ethical questions used to be mostly confined to science fiction, but as we hurtle toward the singularity, they are becoming all too real.
The Future Of Technology Is All Too Human
The idea of approaching a technological singularity is both exciting and scary. While the prospects of technologies that are hundreds of times more powerful than what we have today will open up completely new possibilities, there are also inherent dangers. How autonomous should we allow robots to become? Which genes are safe to edit and which are not?
We will also have to change how we view work. Much like in the industrial revolution when machines replaced physical labor, new technologies are now replacing cognitive tasks. Humans, therefore, will have to become more adept at things that machines can’t do, namely dealing with other humans, and social skills will trump cognitive skills in the marketplace.