211
A LOOK AT THE FUTURE
March 26, 1999
Often I get so busy trying to figure out what will happen next week or next
year that I fail to think about where the world, and by implication our
industry, is headed long range. One of the most thought provoking
presentations at the 1999 Spring Seybold Publishing Conference in Boston was
made by Ray Kurzweil, author of the "The Age of Spiritual Machines."
Kurzweil provided a daunting view of where technology is headed in the next
30 years. I'm 62 and often marvel at the changes that have taken place since
I was in my 20's. However, now I've decided I definitely want to stay around
until I'm at least in my 90's in order to observe the totally changed world
of 2030. For those of you who are younger and plan to be working in some
photography related career for the next 20 or 30 years, it is well worth
considering what the world is likely to be like in that very short period
time.
In order to understand why Kurzweil's predictions are likely to occur, we need
to examine a few scientific principles that Kurzweil outlines at the
beginning of his lecture and book.
Moore's Law
Gordon Moore, an inventor of the integrated circuit and then Chairman of
Intel, noted in 1965 that the area of a transistor was reduced by half
approximately every 24 months. Thus, every two years you can put twice the
number of transistors on a chip and double its speed.
Law of Time and Chaos
In any process, the time interval between salient events (i.e. events that
change the nature of the process, or significantly affect the future of the
process) expand or contract along with the amount of chaos. The more chaos
and disorder the longer it takes for change to come about.
Rate of Accelerating Returns
As a process evolves, unworkable theories are tested and rejected. More and
more order is brought into the system and time between changes gets shorter
and shorter. This is defined by Kurzweil as the Law of Accelerating Returns.
"As order exponentially increases, time exponentially speeds up (i.e. the
interval between salient events grows shorter)."
With Moore's Law and the Rate of Accelerating Returns we find that we have an
exponential rate of improvement in speed and processing capacity in
computing. At the beginning of the 20th century we were doubling computer
power every three years. By 1950-1960 we were doubling computer power every
two years and since 1985 we have been doubling computer power every 12
months.
Chip manufacturers have confidence they can continue this doubling for
another 20 years. At that time transistor insulators will be just a few
atoms thick and the conventional approach to shrinking circuits will no
longer work.
Exponential growth can be very decieving, but it looks like it will continue
through the 21st century. To better illustrate exponential growth their is a
story about the inventor of chess and the Emperor of China. When the Emperor
asked the inventor what he would like in payment the inventor modestly said,
"a single grain of rice on the first square of the chess board, and double
that on each succeeding square." On the 32nd square the amount equalled 4
billion grains of rice and by the 64th square it equalled 18 million trillion
grains of rice, which would require rice paddies covering twice the surface
area (including oceans) of the earth.
Computing Power of the Human Brain
To get an idea of where this technological explosion is headed consider the human brain. The
brain has approximately 100 billion neurons and an estimated 1000 connections
to each neuron. This makes 100 trillion connecting points, each capable of
simultaneous calculations. However, given the slow speed of carbon-based
reactions each connection point is only capable of about 200 calculations per
second for a combined total of 20 million billion calculations per second.
In 1997 $2,000 of neural computer chips using only modest parallel processing
could perform 2 billion connection calculations per second. This capacity
will double every 12 months or 23 doublings by 2020. At that time $2,000
worth of computer chips (in 1997 dollars) will be able to do 20 million
billion calculations per second equivalent to the human brain's calculating
capacity.
Memory in the brain is approximately 100 trillion synapse strenths or about a
million billion bits. In 1998 a billion bits of RAM cost $200. The capacity
of memory circuits has been doubling every 18 months. Thus by 2023 a million
billion bits of memory, equivalent to all the memory in a single human brain,
will cost $1,000.
Thus, it will be possible to produce personal computers with the capacity for
human thinking for about $3,000. Supercomputers with this calculating
capacity will appear around 2010.
Kurzweil claims that it is a conservative statement to say that by 2029 we
will be able to look inside someones brain and create a huge database that
copies every single detail relevant to human thinking. We will be able to
copy all the connections, all the neuro transmitter strengths, all the
synaptic cleffs, and every local process that bears on our ability to
interact with the world and process information.
Interestingly, the cost of production of the integrated circuit remains
fairly constant so this increased speed and capacity is achieved and no
additional costs. However, the silicon equivalent of the carbon-based brain
will operate a billion times faster than the human brain.
Cutting Edge Today
What are some of the cutting edge technologies that are available right now?
IBM's Deep Blue defeated Gary Kasparov in a championship chess match in 1997.
Deep Blue can do 10 trillion calculations per second. This is only 2,000 times
slower than the human brain. With exponential growth they won't have to do much
to catch up with "brain speed".
Using MRI scanning technology we can look inside the human brain and see
individual nerve cells firing. The speed, the resolution and the band width
of these non-invasive scanning technologies is accelerating as we get faster
computers.
Parkinson's disease is caused when a small locus of maybe 10,000 neurons
in two tiny regions of the brain get scrambled. In Europe they are doing
neural implants to relieve the symptoms of this disease. The experimental
implants can be switched on and off at will.
A similar chip is being developed to suppress the tremors associated with
cerebral palsey and multiple sclerosis. A electronde will be implanted in
the ventral lateral thalamus.
Cochlear implants together with electronic speech processors can perform
frequency analysis similar to that performed by the inner ear and enable 10%
of formerly deaf persons to hear and understand voices well enough to hold
conversations.
Kurzweil tells of a deaf friend who has a cochlear neural
implant that replaces early auditary circuits of his nervous system. This
enables him to talk on the phone with Kurzweil.
An experimental retina implant uses a solar powered computer that
communicates with the optic nerve. The user wears special glasses with tiny
television cameras that communicate to the implanted computer by laser
signals.
LOOKING AHEAD
By 2009
Personal computers will be commonly embedded in wristwatches, rings,
earrings, and other body ornaments. They provide communication facilities
similar to cellular phones, pagers, web surfers, as well as monitor body
functions and provide automated identity and directions for navigation.
Many books, music albums, movies and other digital files will be typically
distributed through wireless networks and have no physical objects associated
with them. Computer displays have all the quality of paper - high
resolution, high contrast, large viewing angle and no flicker. Books,
magazines and newspapers are routinely read on displays. Eye glasses have
tiny lasers built in that project images directly on the users retinas.
The majority of text is created using continuous speech recognition software
which is more accurate than human transcribers. Students of all ages
typically own a thin tablet like computer, weighing under a pound, that has a
display suitable for reading. They interact with this device by voice and a
pointing device like a pencil.
Most portable computers will not have keyboards. Short-distance wireless
technology will be used to communicate between components such as pointing
devices, microphones, displays, printers.
Documents created at this time routinely include embedded
moving images and sound. Learning materials are accessed through wireless
communication. Students
learn math and reading from interactive learning devices. Synthetic voices
sound fully human. Schools increasingly rely on software approaches to teaching
all subject matter. The role of human teachers is primarily in motivation,
psychological well-being and socialization.
Print-to-speech reading devices for the blind are inexpensive. Deaf persons
commonly use speech-to-text listening machines which display real-time
transcription.
Translating telephone technology (TTT) is commonly used for many language pairs.
Telephone communications is primarily wireless, and routinely includes
high-resolution moving images. Users can instantly download books,
magazines, newspapers, TV, radio and movies to their highly portable personal
communication devices.
The distribution of all types of written documents from articles to books
typically does not involve paper and ink.
At least half of all business transactions will be conducted on-line.
Intelligent assistants which combine continuous speech recognition,
natural-language understanding, problem solving and animated personalities
routinely assist with finding information, answering questions, and
conducting transactions.
By 2019
Computer will be largly invisible, but embedded everywhere - in walls,
tables, chairs, desks, clothing, jewelry and bodies. People will routinely
use three-dimensional displays built into glasses or contact lenses. This
technology projects images directly onto the human retina, exceeds the
resolution of human vision and is widely used regardless of visual
impairment.
Most interaction with computing is through gestures using hands, fingers and
facial expressions and through two way spoken communications. Connectivity
is wireless and cables have largely disappeared.
Hand-held displays are less common. Direct-eye displays where text is projected
into the ever present virtual environment comes into
greater use. Paper books and documents are rarely used or accessed. Most
learning is accomplished using intelligent software-based simulated teachers.
When human teachers are used they are often not in the local vicinity of the
student and are more mentors and counselors than sources of knowledge.
Blind persons routinely use eyeglass-mounted reading-navigation systems.
Automated reading-navigation assistants communicate to blind users through
both speech and tactile indicators. Paraplegic and some quadriplegic persons
routinely walk and climb stairs through a combination of computer-controlled
nerve stimulation and exoskeletal robotic devices. Generally, disabilities
such as blindness, deafness, and paraplegia are not noticeable and are not
regarded as significant.
Phone calls routinely include high-resolution three-dimensional images
projected through the direct-eye displays and auditory lenses. Reading
books, magazines, newspapers, other documents, listening to music or watching
three-dimensional moving images are accomplished through the communication
Web and do not require any equipment, devices or objects that are not worn or
implanted.
Automated driving systems have been found to be highly reliable and have now
been installed in nearly all roads. Humans are still allowed to drive on
local roads but automated driving systems are always engaged and ready to
take control when necessary to prevent accidents.
By 2019 the computational capacity of a $4,000 computing device (in 1999 dollars)
will be approximately equal the processing power of the human
brain. However, at 200 calculations per second the brain is 10 million times "slower" than
an electronic circuit. So by increasing the number of parallel processors it will
will be easy to the processing capacity of humans.
By 2029
$1,000 worth of computation devices (in 1999 dollars) has the computing
capacity of approximately 1,000 human brains. If you combine the total
computing capacity of all humans on earth with the computing capacity humans
have caused to be built more than 99 percent is non-human.
Much of the nonhuman computing will be conducted on massively parallel neural
nets which are based on reverse engineering of the human brain. Many
-- but less than a majority -- of the specialized regions of the brain have
been "decoded" and their massively parallel algorithms will have been
deciphered.
Displays are now implanted in the eyes. They also act as cameras to capture
visual images and thus function as both input and output devices. A range of
neural implants has become widely available, not just for disabled people.
Many will be using them to enhance visual and auditory perception as well as
interpretation, memory and reasoning.
Human learning is primarily accomplished using virtual teachers and is
enhanced by widely available neural implants. These implants improve memory
and perception, but are not yet possible to download knowledge directly.
Learning still requires time-consuming human experience and study.
There is almost no human employment in production, agriculture and
transportation. The largest profession is education.
A sharp division no longer exists between the human world and the machine
world. Human cognition is being ported to machines and many machines have
personalities, skill and knowledge bases derived from the reverse engineering
of human intelligence.
Cybernetic artists in all of the arts - musical, visual, literary, virtual
experience, and all others - no longer need to associate themselves with
humans or oganizations that include humans. Many of the leading artists are
machines.
Life expectancy of humans will now be around 120 years.
These are just some of the changes that will take place. In his book, which
I would urge everyone to read, Kurzweil makes a convincing case for why this
is a conservative level of change that might be anticipated, not some wild,
far out, futurist theory that will never come to pass.
Life Cycle of a Technology
Kurzweil also provides an interesting analysis of the Life Cycle of a
Technologies that raises some interesting questions about certain
technologies on which at the present time we greatly depend. He identifies
seven stages in the life cycle of most technologies. They are:
1 - Precursor stage - dreamers imagine a technology, but do not have the
means to make it a reality
2 - Invention - a brief stage where an inventor blends curiosity, scientific
skills and determination to bring a technology to life. Often, initially it
does not have much significance in the world.
3 - Development - the invention is developed for mass use. Often the
inventor and other guardians of the technology try to protect it and channel
its use.
4 - Maturity - technology has a life of its own. It is interwoven in the
fabric of the community and appears to many observers that it will last
forever. This creates an interesting drama when the next stage arrives.
5 - Pretenders - new technology comes along and threatens to eclipse the
established technology. Enthusiasts prematurely predict victory because they
point out many advantages.
Upon more reflection it is noticed that while providing some distinct
benefits, the newer technology is missing some key elements of functionality
or quality.
If it fails to disloge the established order, the technology conversatives
take this as evidence that the original technology will indeed live forever.
This victory is, however, usually short lived. The new technology finds
ways to improve those weak areas of functionality and the old technology is
driven into obsolescence.
6 - Obsolescence - the old technology lives out its senior years in gradual
decline. This stage usually comprises 5% to 10% of the life cycle.
7 - Antiquity - the technology lingers, but more for historical and artistic
values than for any practical everyday use. Some examples are: horse &
buggy, harpsichord, manual typewriter, electro-mechanical calculator, vinyl
records.
The most profound role of computers is in the communications medium. As
commercial photographers our primarly role is communications.
With this analysis in mind consider two technologies related to our industry
-- the printed book and the recording of images on film.
The following are my ideas, not Kurzweil's, so don't blame him if my logic is
faulty.
The book was invented 400 years ago, a mature technology that is now at the
stage where software-based "virtual" books enter as pretenders. Current
software books are of low resolution, and have problems with flicker,
contrast, viewing angle, battery life, and they do not have the visual
qualities of paper and ink. However, they also have advantages in low weight
compared to physical books, and improved search capabilities particularly for
technical materials. Technological improvements will enable ""virtual" books
to overcome their current handicaps.
Because the book has been around for 400 years the obsolescence phase (which
I think we will enter soon) may be 20 to 40 years. Despite the use of the
computer for transmitting documents paper usage has been going up because we
don't like to read on the screen. But, when the display is as good as paper
we will use less paper. Kurzweil says overall paper usage will begin to fall
during the first decade of the 21st Century. That means fewer pictures
appearing in printed form. (Fortunately, there will however, probably be many
more pictures used in digital form.)
Film was invented about 100 years ago. Clearly digital capture and storage
is here and film is in its obsolescence phase. If statistical averages are
correct that phase should last 5 to 10 years. We may already be well into
that period. As we look at the improvements that have been made in digital
storage in the past ten years and consider the accelerating pace of these
improvements over the next decade it is not hard to see how film use may soon
disappear.
As the web becomes faster and more ubiquitous there will be less demand for
still images and more demand for motion. Theater projection quality may
become less important as the screens on which the images are viewed become
more personal. When the screen size is the retina, you're not going
to need a whole lot of data to achieve maximum resolution.