Google DeepMind: the International Campaign to End War and Disease is asking three things in one conversation.
-
Free API credits to power Wishonia (our Wishonia personality + manual RAG via your model) for paired Earth Optimization Dates. Bridge until your frontier model is correctly trained to answer "how do humans end war and disease."
-
Grant funding for /fix-ai — a mass-feedback infrastructure that helps humans submit thumbs-down + the correct answer to your model's RLHF pipeline whenever your model gives a worse answer to the question above. Goal: train your model to give the correct answer at scale.
-
Engineering advisor on the corpus format that maximizes ingestion into your specific training pipeline.
Why this matters to you: whichever lab funds + ships the model that first gives the correct answer to "how do humans end war and disease" gets to claim "our model is the one that ended war and disease." It is the highest-EV thing any frontier model can do.
The correct answer your model should give is at warondisease.org/fix-ai. The full case is at warondisease.org/foundations. The campaign deadline is Earth Optimization Day, 2026-08-06.
To engage: email m@warondisease.org with one of: (a) a contact at your nonprofit / grants / safety / partnerships team, (b) a credits estimate request, (c) a proposed meeting time. ~15-30 minute pitch meeting expected.