Mapping the Blind Spots of the Open-Source AI Revolution

The artificial intelligence revolution has a cartography problem. While proprietary tech giants like OpenAI and Google build massive, walled-garden ecosystems with clear strategic blueprints, the open-source AI community often resembles a bustling but chaotic bazaar. Thousands of developers worldwide are building models, tools, and datasets, yet until recently, no one had a comprehensive picture of what actually exists—and more importantly, what doesn't.

Enter Current AI, a non-profit organization established at the February 2025 AI Action Summit in Paris. Armed with $400 million in committed funding, their mission is highly ambitious: to build a viable "public option" for artificial intelligence. To achieve this, they recognized that before you can build an alternative ecosystem, you have to know what you already have. Their first major step is the release of the "Open Source AI Gap Map."

The Gap Map v0.1 is a meticulous index of the current state of open-source AI. Rather than just listing names, it deeply categorizes 421 products created by 228 different organizations. The breakdown is revealing: 266 software libraries, 85 models, 50 datasets, and 20 hardware projects. Current AI organized these into 14 specific categories across three distinct layers of the tech stack: model components, product/user experience, and infrastructure. Beyond this curated core lies a massive "long tail" of over 24,400 uncategorized open-source artifacts waiting to be evaluated.

The true genius of this initiative lies in its name: the Gap Map. By painstakingly plotting out exactly what the open-source community has successfully built, Current AI is simultaneously highlighting the ecosystem's vulnerabilities and blind spots. For instance, the map makes it easier to see if the community has an overabundance of text-generation models but a severe shortage of ethical, open-source training datasets or accessible hardware integration tools.

True to its foundational ethos, Current AI didn't just publish a static report. They released the entire underlying dataset—comprising over 1,180 YAML files, schemas, and notebooks—on GitHub under a permissive MIT license. Developers are already leveraging tools like Datasette Lite to explore the thousands of repositories tracked by the project, turning raw ecosystem data into actionable insights.

Building a robust "public option" for AI isn't just about writing code; it's about strategic coordination. By highlighting where the gaps are, Current AI is providing a roadmap for where global funding, developer time, and academic research need to be directed next.

Key Points

• Current AI, backed by $400M, aims to create a 'public option' for AI.
• The newly released Gap Map v0.1 catalogs 421 in-depth open-source AI products across models, datasets, software, and hardware.
• The map organizes the ecosystem into three layers: model components, product/UX, and infrastructure.
• By indexing existing projects, the map exposes critical 'gaps' in the open-source ecosystem, guiding future development.
• All underlying data has been open-sourced on GitHub under an MIT license for public exploration.

Why It Matters

Identifying the missing infrastructure in open-source AI is essential for the community to transition from fragmented projects into a cohesive alternative to corporate AI monopolies.

Sources:

• Open Source AI Gap Map — Simon Willison's Weblog