November 5, 2005

• The Future

  I thought this was a really cool read. I am thinking of buying his book.

  Human 2.0 By Ray Kurzweil
  October 25, 2005

  We are making exponential progress in every type of information technology.
  Moreover, virtually all technologies are becoming information technologies.
  We can reliably predict that in the not too distant future we will reach
  what is known as "The Singularity".

  This is a time when the pace of technological change will be so rapid and
  its impact so deep that human life will be irreversibly transformed. We will
  be able to reprogram our biology, and ultimately transcend it. The result
  will be an intimate merger between ourselves and the technology we create.

  The evidence for this ubiquitous exponential growth is abundant. In my new
  book, The Singularity is Near, I have graphs from a variety of fields,
  including communications, the internet, brain scanning and biological
  technologies, that reveal exponential progress. Broadly, my models show that
  we are doubling the rate of technical innovation every decade. Throughout
  the 20th century, the rate of progress gradually picked up speed. By the end
  of the century, the rate was such that the sum total of the century's
  achievements was equivalent to about 20 years of progress at the 2000 rate.

  Growth in information technology is particularly rapid: we're doubling its
  power, as measured by price-performance, bandwidth, capacity and many other
  measures, every year or so. That's a factor of 1000 in 10 years, a million
  in 20 years, and a billion in 30 years, although a slow, second level of
  exponential growth means that a billion-fold improvement takes only about a
  quarter of a century. The exponential growth of computing goes back over a
  century and covers five main paradigms:

  . Electromechanical computing as used in the 1890 US census.

  . Relay-based computing as used to crack Nazi cryptography in the early
  1940s.

  . Vacuum-tube-based computing as used by CBS to predict the election of
  Dwight Eisenhower as US president in 1952.

  . Discrete-transistor-based computing as used in the first space launches in
  the 1960s.

  . Computing based on integrated circuits, invented in 1958 and applied to
  mainstream computing from the late 1960s.

  Each time it became apparent that one paradigm was about to run out of
  steam, the realisation resulted in research pressure to create the next
  paradigm. Today we have more than a decade left in the paradigm of shrinking
  transistors on an integrated circuit, but there has already been enormous
  progress in creating the sixth main computing paradigm of three-dimensional
  molecular computing, using carbon nanotubes for example. As another example,
  it took us 14 years to sequence the genome of HIV; SARS took only 31 days.

  The result is that we can reliably predict such measures as
  price-performance and capacity of a broad variety of information
  technologies. There are many things we cannot dependably anticipate. Our
  inability to make reliable predictions applies to any specific project. But
  the overall capabilities of information technology in each field can be
  projected. I have been making predictions of this type for more than 20
  years.

  We see examples in other areas of science of very smooth and reliable
  outcomes resulting from the interaction of a great many unpredictable
  events. Consider that predicting the path of a single molecule in a gas is
  essentially impossible, but predicting the properties of the entire gas -
  comprised of a great many chaotically interacting molecules - can be done
  very reliably through the laws of thermodynamics. Analogously, it is not
  possible to reliably predict the results of a specific project or company,
  but the overall capabilities of information technology, comprised of many
  chaotic activities, can nonetheless be dependably anticipated through what I
  call the law of accelerating returns.

  Between 2000 and 2014 we'll make 20 years of progress at 2000 rates,
  equivalent to the entire 20th century. And then we'll do the same again in
  only seven years. To express this another way, we won't experience 100 years
  of technological advance in the 21st century; we will witness in the order
  of 20,000 years of progress when measured by the rate of progress in 2000,
  or about 1000 times that in the 20th century.

  Ultimately, we will merge with our technology. As we get to the 2030s, the
  non-biological portion of our intelligence will predominate. By the 2040s it
  will be billions of times more capable than the biological part.

  Above all, information technologies will grow at an explosive rate. And
  information technology is the technology that we need to consider.
  Ultimately, everything of value will become an information technology: our
  biology, our thoughts and thinking processes, manufacturing and many other
  fields. As one example, nanotechnology-based manufacturing will enable us to
  apply computerised techniques to automatically assemble complex products at
  the molecular level.

  This will mean that by the mid 2020s we will be able to meet our energy
  needs using very inexpensive nanotechnology-based solar panels that will
  capture the energy in 0.03 per cent of the sunlight that falls on the Earth,
  which is all we need to meet our projected energy needs in 2030.

  A common objection is that there must be limits to exponential growth, as in
  the example of rabbits in Australia. The answer is that there are, but
  they're not very limiting. By 2020, $1000 will purchase 1016 calculations
  per second (cps) of computing (compared with about 109 cps today), which is
  the level I estimate is required to functionally simulate the human brain.
  Another few decades on, and we will be able to build more optimal computing
  systems. For example, one cubic inch of nanotube circuitry would be about
  100 million times more powerful than the human brain. The ultimate
  1-kilogram computer - about the weight of a laptop today - which I envision
  late in this century, could provide 1042 cps, about 10 quadrillion (1016)
  times more powerful than all human brains put together today.

  And that's if we restrict the computer to functioning at a cold temperature.
  If we find a way to let it get hot, we could improve that by a factor of
  another 100 million. And we'll devote more than 1 kilogram of matter to
  computing. We'll use a significant portion of the matter and energy in our
  vicinity as a computing substrate.

  Our growing mastery of information processes means the 21st century will be
  characterised by three great technology revolutions. We are now in the early
  stages of the "G" revolution (genetics, or biotechnology). Biotechnology is
  providing the means to change your genes: not just designer babies but
  designer baby boomers. But perfecting our biology will only get us so far.

  Biology will never be able to match what we will be capable of engineering,
  now that we are gaining a deep understanding of biology's principles of
  operation. That will bring us to the "N" or nanotechnology revolution, which
  will achieve maturity in the 2020s. There are already early impressive
  experiments. A biped nanorobot created by Nadrian Seeman and William
  Sherman, of New York University, can walk on legs only 10 nanometres long,
  demonstrating the ability of nanoscale machines to execute precise
  manoeuvres.

  Microchips of Bedford, Massachusetts, has developed a computerised device
  that, when implanted under the skin, delivers precise mixtures of medicines
  from hundreds of nanoscale wells inside it. There are many other examples.

  By the 2020s, nanotechnology will enable us to create almost any physical
  product we want from inexpensive materials, using information processes. We
  will be able to go beyond the limits of biology, and replace your current
  "human body version 1.0" with a dramatically upgraded version 2.0, providing
  radical life extension. The "killer app" of nanotechnology is "nanobots",
  blood-cell-sized robots that can travel in the bloodstream destroying
  pathogens, removing debris, correcting errors in DNA and reversing ageing
  processes.

  We're already in the early stages of augmenting and replacing each of our
  organs, even portions of our brains with neural implants, the most recent
  versions of which allow patients to download new software to their implants.

  The most profound transformation will be "R" for the robotics revolution,
  which really refers to "strong" AI, or artificial intelligence at the human
  level. Hundreds of applications of "narrow AI" - machine intelligence that
  equals or exceeds human intelligence for specific tasks - already permeate
  our infrastructure. Every time you send an email or make a mobile phone
  call, intelligent algorithms route the information. AI programs diagnose
  electrocardiograms with an accuracy rivalling that of doctors, evaluate
  medical images, fly and land aircraft, guide intelligent autonomous weapons,
  make automated investment decisions for over $1 trillion of funds, and guide
  industrial processes. A couple of decades ago these were all research
  projects.

  With regard to strong AI, we'll have both the hardware and software to
  recreate human intelligence by the end of the 2020s. We'll be able to
  improve these methods and harness the speed, memory and knowledge-sharing
  ability of machines.

  Ultimately, we will merge with our technology. This will begin with nanobots
  in our bodies and brains. The nanobots will keep us healthy, provide
  full-immersion virtual reality from within the nervous system, provide
  direct brain-to-brain communication over the internet and greatly expand
  human intelligence. But keep in mind that non-biological intelligence is
  doubling each year, whereas our biological intelligence is essentially
  fixed.

  As we reach the 2030s, the non-biological portion of our intelligence will
  predominate. By the mid 2040s, the non-biological portion of our
  intelligence will be billions of times more capable than the biological
  portion. Non-biological intelligence will have access to its own design and
  will be able to improve itself in an increasingly rapid redesign cycle.

  This is not a utopian vision: the GNR (genetics, nanotechnology and
  robotics) technologies each have perils to match their promises. The danger
  of a bioengineered pathological virus is already with us. Self-replication
  will ultimately be feasible in non-biological nanotechnology-based systems
  as well, which will introduce its own dangers. In short, the answer is not
  relinquishment. Any attempt to proscribe such technologies will not only
  deprive human society of profound benefits, but will drive such technologies
  underground, which would make the dangers worse.

  We won't experience 100 years of technological advance in the 21st century;
  we will witness in the order of 20,000 years of progress when measured by
  the rate of progress in 2000, or about 1000 times that achieved in the 20th
  century. Some commentators have questioned whether we would still be human
  after such dramatic changes. These observers may define the concept of human
  as being based on our limitations, but I prefer to define us as the species
  that seeks - and succeeds - in going beyond our limitations.

  Because our ability to increase our horizons is expanding exponentially
  rather than linearly, we can anticipate a dramatic century of accelerating
  change ahead.

  Intuitive view

  In 2003, Time magazine organised a "Future of Life" conference celebrating
  the 50th anniversary of Watson and Crick's discovery of the structure of
  DNA. All speakers, myself included, were asked what we thought the next 50
  years would bring. Most of the predictions were short-sighted.

  James Watson's prediction was that in 50 years, we'll have drugs that allow
  us to eat as much as we want without gaining weight. "Fifty years?" I
  replied. We've already demonstrated it in mice and human drugs using the
  relevant techniques that are in development. We can expect them in five to
  10 years, not 50.

  The mistake that Watson and virtually every other presenter made was to use
  the progress of the past 50 years as a model for the next half-century. I
  describe this way of looking at the future as the "intuitive linear" view:
  people assume the current rate of progress will continue. But technological
  change is not linear but exponential. You can examine the data in different
  ways, and for a variety of technologies, from electronic to biological. You
  can analyse the implications, from the sum of human knowledge to the size of
  the economy. However you measure it, the exponential acceleration of
  progress and growth applies.

  Understanding exponential progress is key to understanding future trends.
  Over the long term, exponential growth produces change on a scale
  dramatically different from linear growth. Consider that in 1990, the human
  genome project was widely regarded as controversial. In 1989, we sequenced
  only one-thousandth of the genome. But from 1990 onwards the amount of
  genetic data sequenced doubled every year - a rate of growth that continues
  today - and the transcription of the human genome was completed in 2003.

  Reverse-engineering the brain

  The most profound transformation will be in "strong" AI, that is, artificial
  intelligence at the human level.

  To re-create the capabilities of the human brain, we need to meet both the
  hardware and software requirements. Achieving the hardware requirement was
  controversial five years ago but is now largely a mainstream view among
  informed observers.

  Supercomputers are already at 100 trillion (1014) calculations per second
  (cps) and will hit 1016 cps near the end of this decade, which is the level
  I estimate is required to functionally simulate the human brain. Several
  supercomputers with 1015 cps are already on the drawing board, with two
  Japanese efforts targeting 1016 cps about the end of the decade. By 2020,
  1016 cps will be available for about $1000. So now the controversy is
  focused on the algorithms.

  To understand the principles of human intelligence we need to
  reverse-engineer the human brain. Here, progress is far greater than most
  people realise. The spatial and temporal resolution of brain scanning is
  progressing at an exponential rate, roughly doubling each year. Scanning
  tools, such as a new system from the University of Pennsylvania, can now see
  individual interneuronal connections and watch them fire in real time.
  Already, we have mathematical models of a couple of dozen regions of the
  brain, including the cerebellum, which comprises more than half the neurons
  in the brain.

  IBM is creating a highly detailed simulation of about 10,000 cortical
  neurons, including tens of millions of connections. The first version will
  simulate electrical activity and a future version will also simulate
  chemical activity. By the mid-2020s, it is conservative to conclude that we
  will have effective models of the whole brain.

  One benefit of a full understanding of the human brain will be a deep
  understanding of ourselves, but the key implication is that it will expand
  the tool kit of techniques we can apply to create artificial intelligence.
  We will then be able to create non-biological systems that match human
  intelligence. These superintelligent computers will be able to do things we
  are not able to do, such as share knowledge and skills at electronic speeds.

  New Scientist

  Futurist, inventor and writer Ray Kurzweil is responsible for innovations
  such as text-tospeech synthesisers and the first musical instrument
  synthesiser. His books include The Age of Intelligent Machines (1990), The
  Age of Spiritual Machines (2000) and his new book, released this month, The
  Singularity is Near: When Humans Transcend Biology, on which this story is
  based.

  www.kurzweiltech.com, kurzweilai.net, singularity.com

  • 4:40 am
  • 1 Comment