THE ASCE DANCE OF SCIENCE AND TECHNOLOGY
547
Each dot plotting the exponential growth represents a project or a
product. So each dot represents a human story, drama, venture capital,
IPO, competition, bankruptcy. You'd think, therefore, that the trend
would be very erratic. Maybe the general trend would be there, but it
wouldn't make such a smooth, predictable curve. It's remarkable just
how predictable these curves are. Communications are just as profound
as computation. There are twenty different ways to look at it, but all of
these trends move with exponential growth . And if you get enough data,
you can actually see the cascade of S-curves . That's how technology
evolves. A particular approach-a technology, a paradigm-starts out
slowly and explodes exponentially, but then reaches its limit. But then
another paradigm takes over and continues the exponential growth. So
what keeps the exponential growth going when a particular paradigm
hits its asymptote is innovation-paradigm shift.
The book I wrote in the 1980s predicted that the Internet would be
a ubiquitous phenomenon by the mid-1990S. In the mid-1980s we went
from twenty thousand nodes to forty thousand in one year, and then
doubled it to eighty thousand. But that's a very small number. Only a
few thousand scientists were using it. Nobody had heard of the Internet.
But by the mid-T990S it would reach twenty million, then double to
forty million and eighty million, and then it would be on people's radar
screens. We don't experience technology in the exponential domain; we
live in a linear world. So in most people's perception, the Internet just
came out of nowhere in the mid-1990S. But if you look to the expo–
nential trend, you could see it coming.
Another very important trend is miniaturization. This is the decrease
in transistor size, which is a reflection of Moore's Law, but it is also true
of mechanical devices. We're now at the stage of technology called
MEMS (Micro Electronic Mechanical Systems), which is using the same
technology that we use to build integrated circuits to build tiny
machines. There's actually an intersection of this technology, MEMS,
with electronics and biology. There are four major conferences on things
called bioMEMS, devices the size of blood cells which will actually go
inside the blood stream for medical purposes, built using this type of
technology. Not quite nanotechnology yet, these are a little bigger than
the multi-nanometer scale. One example is a tiny device you put in pan–
creatic islet cells, where there are little pores that are seven nanometers
in size.
It
lets the insulin out but blocks the antibodies, which are a lit–
tle bigger than the insulin molecules, and lets oxygen in to feed the pan–
creatic islet cells . This has actually cured type one diabetes in rats, and
there's no reason why the same methodology couldn't work in humans.
