Community Owned Contextual Deployment Data
EVIDENCE-BASED DECISION-MAKING
CROSS-FUNCTIONAL COALITIONS
COMMUNITY AGENCY & BENEFIT
Only 9% of Americans believe AI will do more good than harm, yet it is increasingly embedding itself within society, often in the name of social impact.
The "AI for Good" movement promotes spreading frontier models for "societal benefit" despite limited evidence of meaningful impact, low public desire, and self-admitted claims that already-present harms can escalate.
This happens because of a top-down push from Big Tech that's occurring without sufficient evidence or input from end users. This often manifests as a partnership with a "trust based institution" who is able to look past their mass, non-consensual collection of training data, environmental harms, and contracts in defense, surveillance, and other oppressive systems. This an effort to improve their poor public sentiment and penetrate the market at a time of light regulation while circumventing accountability as downstream communities absorb risks.
The Kaleidoscope Project is an attempt to counterbalance this top-down push by researching and facilitating ethically-sourced, community-owned contextual deployment data to strengthen evidence-based decision making, strategic cross-functional coalitions, and develop tools for democratic, transparent governance.
Coming Soon: Our pilot research project is mapping whose work fits into one of our Kaleidoscope categories. Fill out a short survey will help produce a public report at minimum and may connect you with those with shared goals but unique approaches.
The Kaleidoscope Model
While there are many nuanced approaches to "AI for Good", they can generally be categorized into one of 6 groups depending on their structural power. Read more here [link] for more gaps, along with strengths and weaknesses in their general execution.
The current power distribution is primarily concentrated in AI labs, and secondarily in a subset of funders and governments. Builders and researchers sit below them, while end users have the least.
At Kaleidoscope, we advocate for a more equitable power distribution centered on community-owned, contextual-level deployment data that guides how AI is deployed and governed.
For an in-depth overview of how the Kaleidoscope works in practice, including examples, click here. The high-level operational rules include:
Meaningful community engagement and approval early-on in any intervention.
Community ownership over contextual-deployment data (that others pay to access).
AI labs can respond but should not initiate an intervention given their conflicts of interests
Think that this work could be improved? We would love your feedback!