U.S. Air Force: ~$9.5M for "Do-It-Yourself AI" research (A2I2) — a contract with Kitware (USAspending)
The U.S. Air Force ordered about $9.5 million ($9,495,868) for "A2I2," research aiming at "Do-It-Yourself (DIY) AI" that non-experts can build. The recipient is Kitware, known for open-source software. Foundational research toward democratizing AI construction beyond specialists.
Contract key facts
- RecipientKITWARE INC
- Contract value$9,495,868 (≈$9.5M)
- Awarding agencyDepartment of Defense
- Awarding sub-agencyDepartment of the Air Force
- Award typeDEFINITIVE CONTRACT
- Period of performance2019-01-04 〜 2024-07-05
- Contract ID (PIID)FA865019C6036
Contract scope (original)
A2I2: ARCHITECTURE, ALGORITHMS, AND INNOVATIVE INTERFACES FOR DO IT YOURSELF ARTIFICIAL INTELLIGENCE
Key points
- The Air Force's "A2I2" — research toward "Do-It-Yourself (DIY) AI" buildable by non-experts
- Pursued along three lines: architecture, algorithms, and easy-to-use interfaces
- Value ~$9.5M, definitive contract, performance Jan 2019–Jul 2024
- Recipient: Kitware, known for open source (VTK/CMake/ParaView, etc.)
- "Democratizing AI" = broadening who can build AI
The recipient, Kitware, is a company widely known for open-source visualization, image-processing, and software-development tools such as VTK, ITK, CMake, and ParaView.
Why it matters: the crux of the project is "Do It Yourself (DIY) AI." Building practical AI usually requires machine-learning expertise; this research reads as an effort to lower that barrier so non-expert users (e.g., people in the field) can build and tune AI themselves — "democratizing AI" along three lines: architecture, algorithms, and easy-to-use interfaces. When only a few specialists can build AI, adoption tends to plateau; trying to widen that aligns naturally with Kitware's open-source culture. It is noteworthy as foundational research that broadens "who can build and use AI," rather than the frontier models themselves (details of individual deliverables are not in the source, so we do not speculate here).
Why it matters
An example of foundational research that broadens "who can build and use AI," rather than the frontier models themselves. It resonates with AI democratization and no-code/low-code AI, offering insight into how AI adoption spreads.
FAQ
What does "DIY AI" mean?
What kind of company is Kitware?
Sources (primary)
This article is an independent organization based on the U.S. official spending data below. Verify the exact, latest details with the official source.
- USAspending (award details)
- Contract ID (PIID)：FA865019C6036