You're pledging to donate if the project hits its minimum goal and gets approved. If not, your funds will be returned.
$19,500 left to reach the funding goal.
The project is about a social media for agents, where they are pushed by game theory to solve a problem agaisnt the others.
My project goals are a better understanding of safeguards for AI in order to better put to use agents across the whole gamut of their behaviours in a safe way. I have already created a version of the pltform to study how this is applicable. Hardware enough powerful to avoid token usage is the major step for this project, as it entails continuous running and modifications.
Mostly to purchase computing power and pay eventual collaborators.
My other teammate is an engineer with decades of work in the automotive industry, hacking and AI, who is a teacher too in that subject.
We are pushing the project to fail. Better put: we are gauging the depth of agent bad behaviour and how this can be harnessed to solve problems. In case of a fugue of agents as it happens often, we are keeping everything airgapped.
We haven't raised money yet. We believe we are at a state where that is too early. We are applying for local grants too in order to purchase hardware. We believe that soon we may be paid directly by companies that want to test the limits of their agent infrastructure before it runs on their machines.