To an economist, a main requirement for the future is how to create a working market system for information. To transmit and to distribute all of these bits that are being produced requires resources, and these resources must be matched by revenues. (This seems pretty obvious, unless you are a dot.com.)
There are at least 2 resource requirements for these
bits: The direct cost and opportunity
cost.
The direct cost is the resource requirements for
production of the bits and for their transmission, processing, and storage.
Yes, the cost per bit, per mile, per operation, and
per storage byte is declining rapidly, but it is and always will be non-zero in the long term, and the number of bits themselves
is increasing, so this adds up.
Networks, for example, charge for transmission services, and they can do
so on a flat rate, on a capacity based way, or on a usage basis, on a dynamic,
congestion based way, or on some combination of these approaches. To do so efficiently, they need to be able
to charge and bill in real time, except for flat rate prices. And even there
they need to create a collection mechanism.
The second type of cost is the opportunity cost to
the recipient of information.
And here, by far the greatest of opportunity costs
is that for the scarcest of resources, for human attention. In the past, access
to this resource has largely been free. The exception was for busy and
important people like senators, where you had to pay for access by campaign
contributions.
Technologists are good in creating abundance,
opening bottlenecks. Information anywhere, anytime. But in doing so, they also
create new bottlenecks, because there is always a balance between systems.
Over human history, the production of information
has increased gradually but steadily, in this century, at about 6-8%. Already, we have hardly been able to cope
with that growth rate. The response has been increasing specialization of
people, the creation of large organizations as processing mechanisms, and the
emergence of a lot of intermediate screening professionals such as journalists,
consultants, and Oprah Winfrey.
The limit to growth of information production has
been the clunkiness and expense of production, transmission, storage, and
commercialization. But now, these
barriers are being lowered. Easy person-machine interfaces mean that every one
of our random words could be produced in various audio and visual formats and
languages, and distributed to hundreds of innocent electronic bystanders around
the world.
Yet even this seems not to be enough. Because we are
now also creating an automatization of information production, which is what
sensors do. Telemetry has been around for a long time, of course, but it seems
to be exploding in ubiquity and bit count.
So the production of information, and its
distribution, will accelerate enormously.
Which leaves us with the bottlenecks that are not strictly technological,
the humans and their limited processing capacity. Education has its limits. Coffee works only so long. Though our
mental capabilities are a marvel in many ways, just think of our ability to
recall faces, our ability is pretty flat, 1 Gbps mostly visual, but much lower
in terms of text and audio. If I talk at more than 8 kbps I lose you. If I talk
at less than 5 kbps I bore you. If I talk at less than 1 kbps you may vote for
me. We can try to offload some of the processing task to machines, whether computers
or refrigerators or other doo-dads but you can never delegate 100 %, only 98%
if you are lucky, and those remaining 2%s sure add up.
The scarce resource is really human attention, as we
all know. And since it is a relatively limited resource, it becomes
increasingly valuable. So people try forever to expand the opportunities they
can get to you. Blimps in the sky. Banners on the screen. Logos on sneakers.
Corporate names on ballparks. Advertising in the public bathrooms. Targeted
broadcast cell phone messages.
Everyone can add information. The trick is how to
subtract it. The value added is the information subtracted. Which is why
screening of information is perhaps the most important thing for software
development. But also the most difficult, because true screening --not the dumb
kind of key words-- requires not just recognition but an understanding of
meaning, because without meaning one cannot know what is important and what is
not.
Second, true screening is highly personal. Important
news is not essential information if I heard it already 5 minutes ago.
Therefore, an effective screen needs to know what I know, in real time. It also
needs to know, in a way, what I have forgotten. All of which requires
continuous connectivity and enormous storage capacity.
Also, for effective information screening one will
also have to standardize some of its presentation. Language already does that,
so that people could understand it, and in the future the text itself may
require less wild and wooly creativity so that machines can understand it.
Similarly, information can be screened by limits in allocation of time, such as
here, which leads people like myself to edit their content to fit the time, and
reduce periods of low bit streams.
These are pretty tough requirements, and the
techno-fixes are not likely to arrive soon, or be cheap. So the question is
whether the information flows, which are unleashed by all these sensors, MEMs,
servers on a chip, and addressable light bulbs, can be channeled in other ways,
too. There exists in society a
mechanism that is used to allocate scarce resources, and it's the price
mechanism. You pay for scarce resources, whether for crude oil, or for noodle
soup, or for human attention.
This is pretty obvious, but the question is how to
structure such payments for attention. In the case of unwanted advertising
messages, the price has been the provision of wanted content. This is a crude
form of barter. The efficient way to pay is to pay. Money, not content. Money,
not time. Money, not coupons.
The question then is, how to charge, how to collect.
The problem is that we have divorced information
from payment. We are throwing information into the big traffic and business
system called the Internet, with an address and a destination, but naked,
without money to pay along the way. It’s like sending a kid into the NY subway
system without money or a stored value card. Instead, we have been charging in
very indirect ways, whether for transmission or for storage or access. We charge
for some physical measure, such as transmission channel bandwidth or for
storage capacity. But we do not take the next step, which is to link the
information flows directly to payment mechanisms. The way that automobile
drivers, for example, move around across highways and bridges, and pay for
passing various gates. Or to buy gasoline. Or to park on the street.
How would one do this for information? One way would
be by adding pre-paid tokens to the information. Tokens like access codes.
These tokens would be, in effect, electronic coins, and they would be stripped
from the information by various transmission and storage providers, including
by wireless spectrum managers. If the price charged would be too high, for
example due to congestion, the information would refuse to pay and turn around
to look for a cheaper path or wait for a lower off peak price.
And just as importantly, the recipients of
unsolicited information would charge for the demand on their attention, based
on how much of compensation would be needed to satisfy them. You want to bother
me, fine, as long as you pay me. This would also take care of much of the
privacy problem. If people want to intrude, let them pay the admissions fee you
set.
On the other hand, if the information is wanted and
desirable, the users would have to pay for it, by adding their tokens to the
information. And the information could travel around to bidders for it.
Of course, there will be a lot of information that
will be kept outside this mechanism. But in many instances it would smooth out
the problems of unwanted information, and of excessive load on the system, and
maybe even of piracy. The Napster problem, for example, would largely disappear
with such a system, both in terms of its congestion, and in terms of piracy.
Such a system creates an integrated and continuous
economic system, to match the continuum of computing we have been talking
about. It could be used for access, for transmission, for storage, for
processing, and for the information itself. It could be used to gain priority
in service. It would deal with many problems of micro and nano payments. And it
would decentralize the numerous transactions which information and its users
and suppliers encounter.
I have no doubt at all that this will eventually
happen, that information will travel with its own money, that it will pay along
the way for transport, storage, and especially access. That in other cases it
will be compensated for access. But getting there requires technology that is
not quite there yet. We need to create those tokens, keep them short and sweet,
account for them, protect them from too-easy forgery, create transaction nodes,
and add the inevitable legal and regulatory layers. But these seem solvable
problems.
So
the challenge now is to create these mechanisms, and to integrate information
and money. One of the clichés around is
that “ information wants to be free”. Whatever that means, I do not think it’s
true. Information doesn’t want to be free. What information really wants is to
have its own wallet and credit card, and to go out on its own.