146 Chapter 4
computational center, a few dozen prominent computational centers, and a
lower network. Such a structure will allow flexible information accounting
and both operational management of the industry according to territories
and the organization of planning accounts according to topics.”^70
After 1967, CEMI internal documents stop mentioning any network
projects, whether OGAS or EGSVTs. What began as a small laboratory
devoted to a wide-ranging civilian-use network for the management of the
economy became, as a RAND report later called it, an “operational support
agency” for the Gosplan.^71 Today CEMI is remembered for spearheading
optimal planning methods with computerized and mathematical models
in the Soviet socialist economy. As its website proclaims with silent hind-
sight on its early network ambitions, “When the Institute was founded in
1963, its main goal was to elaborate the theory of optimal management of
the economy, applying mathematical methods and the use of computers to
the task of practical planning.”^72
Like most rivalries, the subsequent rivalry between these two peer insti-
tutes, CEMI and IK, developed out of, in Freud’s phrase, a narcissism of
petty differences. After having their original OGAS mission tabled, both
resorted to developing from the bottom up microlevel, factory-level eco-
nomic planning. Even today, CEMI continues to pursue enterprise-level
economic modeling, and IK continues to develop automated systems of
management (ASUs) for individual enterprises. Fedorenko reported hav-
ing improved hundreds of factory-level flow models every decade, and
Glushkov claimed to have established ASUs in Ukraine, St. Petersburg, and
beyond. Despite these successes, Glushkov in 1975 observed that humans
entering “half-truths” were hampering automated control systems so that
“we find ourselves somewhere between confusion and a search for scape-
goats.”^73 The few ASUs that were implemented fell flat, as well, accord-
ing to the émigré mathematical economist Aron Katsenelinboigen, who
reported that ASUs had little to no effect and sometimes even negative
effects due to the expenses of installation. Managers, who were often older
and wary of being replaced, often lacked the capacity to become familiar
with, let alone master, the economic-mathematical methods that the ASU
required.^74 As Glushkov later noted in Pravda, one automatic control sys-
tem was dismantled and sold because it “impartially pointed out manage-
ment’s blunders and omissions.”^75 What began as an alliance in the early
1960s around a network became a rivalry after the 1970s when cybernetic
institutes disagreed over the relevance and proper role of the computer in
economic planning. As these sections illustrate, the tensions resulted not
from the roles of computing networks and information technology but