Artificial intelligence did not grow in a vacuum, and for most of its history it was not funded by industry. For decades the largest patron of AI research in the United States was the military, through DARPA, and the shape of the field, what got tried, what got overpromised, and what got abandoned, tracked the flow of that money closely.
The idea
Research follows funding. AI’s booms and winters, and the directions the field pursued, were driven substantially by DARPA’s decisions about what to pay for.
The Strategic Computing Initiative
From 1983 to 1993, DARPA’s Strategic Computing Initiative funded research into advanced computer hardware and artificial intelligence, directed by the Information Processing Technology Office. It set out to develop machine intelligence within a prescribed ten-year window, and it spanned the whole stack, from chip design and manufacture through computer architecture up to artificial intelligence software.
What it paid for
The initiative backed concrete, ambitious projects. Among them were the Autonomous Land Vehicle, an early attempt at a self-driving vehicle, and the Dynamic Analysis and Replanning Tool, known as DART, used for military logistics planning. The mix is telling: a moonshot at machine autonomy alongside a practical tool that delivered real value.
The scale
The money was substantial. By 1985 the initiative had spent 100 million dollars across 92 projects at 60 institutions, split between industry on one side and universities and government labs on the other. Over its full run, the Department of Defense spent a total of one billion dollars on it.
Why it matters
The artificial intelligence in use today descends in part from work the military paid for when no one else would. The field’s long history of soaring ambition followed by disappointment, the so-called AI winters, is in large measure the history of that funding rising and falling. Understanding where AI came from means following the money, and for a long time the money was DARPA’s. The questions that raises about who steers the technology are the subject of AI governance.
Related Notes
- AI Governance, the modern politics of who directs AI
- Could an LLM Be Conscious?, the philosophical question hanging over the field DARPA set in motion
- Turing and Computability, the theory under the field DARPA funded
- Semiconductor Supply Chains, the hardware the initiative also funded
- ILLIAC IV and Parallel Processing, an earlier ARPA computing bet
- Computing and the U.S. Military, the cluster index
Sources
- “Strategic Computing Initiative,” Wikipedia. https://en.wikipedia.org/wiki/Strategic_Computing_Initiative . Supports the initiative as a DARPA program directed by the Information Processing Technology Office, running from 1983 to 1993 and funding advanced computer hardware and artificial intelligence from chip design through AI software, the Autonomous Land Vehicle and the DART replanning tool as funded projects, the figure of 100 million dollars spent by 1985 across 92 projects at 60 institutions, and the total Department of Defense spend of one billion dollars.