I use AI tools every day — for learning, building, and generally trying to get better at what I do. But the more I learned about the environmental cost of large language models, the harder it became to ignore: every prompt has a footprint, and those footprints add up fast across millions of users.
I didn't want to stop using AI. The upskilling is real, the productivity gains are real, and the technology is genuinely exciting. What I wanted was a way to use it responsibly — to keep growing while giving something back to the planet in proportion to what I was taking.
The idea behind EcoDues is simple: calculate the actual damage your inference usage causes (in dollars, using the social cost of carbon), then donate twice that amount to the climate charities most likely to make a real difference. You keep the productivity. The planet gets a net positive.
This is a solo project built with a lot of curiosity and the help of AI tools themselves — which felt fitting. It's not perfect, and the methodology will keep improving as better data becomes available. That's why it's open source.