Articles /
Why MAUB
AI is everywhere in the talk and rarely where it would actually help. MAUB starts from the other end: needs, and the projects that already meet them.
The MAUB team ·
There is a lot of talk about artificial intelligence, almost always in the same terms: productivity, markets, the race for bigger models. Meanwhile, low-key projects use it to identify plants, spot deforestation, read the labels on our food or estimate emissions that nobody reports. Others would not exist without it: a non-profit can now code in a few weeks the tool it could never have paid a development team to build.
MAUB (Make AI Usage Better) starts from there. The site does three things.
Listing
A directory of AI projects that are useful to society and focused on ecology: AI is part of the tool, or it helped build it. Each entry says what the project does, who runs it and what AI does in it, with a label that states the model, where it runs and who stays in charge. Anyone can suggest a project; every suggestion is reviewed before it is published.
Highlighting
Articles that tell these projects' stories at greater length: how they started, what works, what gets stuck, what can be learned from them. They try to show what a sensible use of technology can achieve.
Building
Anyone can describe a need that nothing meets yet, in their town or their organisation. Others support it. The needs with the most support are looked into and, where possible, built: by pointing to an existing project, or by building what is missing.
The criteria
A project is listed here if it serves the common good and is honest about how it works. In practice we look at five things, set out in the manifesto: real usefulness, transparency about the use of AI, frugality, open code and data, and the room left for the community.
The site tries to live up to them too: no advertising, no trackers, nothing that depends on JavaScript, no passwords, and an API for anyone who wants to reuse the directory.