attended classes sponsored by AT&T so that he would get with its pro-
gram. But the more Baran saw, the more convinced he was. And in a final
push, the Defense Department offered simply to pay AT&T to build the sys-
tem. The government promised no risk; it wanted only cooperation. But
even here, AT&T balked. As recounted in John Naughton's _A_Brief_History_
_of_the_Future_:[3-22]

____ [AT&T's] views were once memorably summarised in an exasperated out-
____ burst from AT&T's Jack Osterman after a long discussion with Baran.
____ 'First,' he said, 'it can't possibly work, and if it did, damned if we are going
____ to allow the creation of a competitor to ourselves.'[3-23]

"Allow." Here is the essence of the AT&T design, supported by the state-
sanctioned monopoly. In "defend[ing] the monopoly,"[3-24] it reserved to itself
the right to decide what telecommunications would be "allowed." As Baran
put it, AT&T "didn't want anybody in their vicarage."[3-25] It controlled the
wires; nothing but its technology could be attached, and no other system of
telecommunications would be permitted. One company, through one re-
search lab, with its vision of how communications should occur, decided.
Innovation here, for this crucial aspect of modern economic life, was as this
single organization would decide.

Now again, the point is not that AT&T was evil. Indeed, quite the con-
trary. We get nowhere in understanding how systems of innovation work
when we personify organizations and imagine them responsible for social
goals. AT&T had an obligation to its stockholders; it had an obligation to
the government to assure consistent quality service. It was simply acting to
assure that it met both of these obligations -- maximizing its profits for its
shareholders while meeting its obligations to the government.

But what's good for AT&T is not necessarily good for America. What
AT&T was doing may well have made sense for it; its vision of tele-
communications may well have made sense for the interests it understood
itself to be serving. But AT&T's vision of what a telecommunications service
should be is not necessarily what a telecommunications service should be.
There is a possible -- and in this case actual -- conflict between the interests
of a centralized controller of innovation and the interest in innovation
generally.

Here the conflict was plain. If the Defense Department built a tele-
communications system based on packets rather than circuits, then the effi-
ciency of that system could in theory be much greater. When you're on a

[[32]]

p031 _ -chap- _ toc-1 _ p032w _ toc-2 _ +chap+ _ p033

v?