ARPANET was a public research investment in networking before there was a commercial Internet market, funding universities, contractors, and shared infrastructure whose spillovers became global.
Douglas Engelbart's NLS was financed by ARPA, NASA, and Air Force research money before personal computing existed, creating a laboratory for the mouse, hypertext, collaborative editing, and interactive work.
Project MAC was a public R&D investment in interactive computing that funded time-sharing, a new computer-science institution, and research that later fed networking and operating systems.
The Apollo Guidance Computer was a concentrated public investment in embedded computing, integrated circuits, and ultra-reliable software whose spillovers reached far beyond the Moon program.
ARPA was created after Sputnik to finance advanced projects that cut across military-service boundaries and exceeded ordinary procurement horizons. Its early hundreds of millions created a model for high-risk public technology investment.
SAGE consumed government-scale capital to build a continent-wide real-time air-defense network. Its direct strategic return was hard to measure, but its software, interfaces, memory systems, and industrial supplier base helped create modern computing.
Whirlwind cost the U.S. government about $7.5 million, yet its return reached far beyond one MIT computer: real-time computation, interactive displays, core memory, and the technical lineage that led to SAGE and DEC.