The Adaptive Layer

Raising $20,000active

Exploring privacy-preserving infrastructure that helps AI-enabled systems adapt to human needs while preserving agency, accessibility, and meaningful participation.

Project Details

People shouldn’t have to repeatedly explain how they should be served. We’ve built excellent infrastructure for proving identity, but very little infrastructure for communicating human context. I propose the Adaptive Layer: a standards-based, open-source framework for securely and selectively communicating user-controlled adaptive attributes through trusted digital credentials.

Rather than communicating identity alone, trusted credentials could optionally communicate selected adaptive attributes with explicit user consent. The objective is not to disclose more information. It’s to disclose only what is necessary to create a better interaction.

This project is currently me, myself, and I exploring use cases and potential collaborations with other like-minded folks and organizations. I’d like to bring an IRL experience to life where people can feel what adaptive environments could become, paired with an open-source framework that artists, designers, technologists, museums, cities, healthcare systems, schools, and public spaces could build on over time.

Theory of Impact

As AI becomes embedded in healthcare, employment, education, finance, transportation, and public serves, more human decisions and opportunities will be mediated by systems that are optimized around standardized users. If those systems can’t recognize human variation without collecting extensive personal data, they may deepen exclusion, increase dependence on opaque institutions, and reduce people’s ability to participate meaningfully in society.

The Adaptive Layer would explore whether trusted, user-controlled attributes can help AI-enabled systems respond to individual needs while preserving privacy, consent, and agency. My theory is that societies will be better able to oversee and govern increasingly capable AI if people remain included, informed, and capable of participating in the institutions shaped by it. This would not solve AI existential risk by itself, but it may strengthen the human resilience and pluralism needed to recognize, challenge, and respond to dangerous uses of AI.

I also believe this could have preventative value beyond accessibility alone. By reducing exclusion and making participation easier and more dignified, adaptive systems may strengthen belonging, agency, and access to purposeful activity, helping address some of the conditions associated with deaths of despair. I see that as a hypothesis worth testing carefully, not a guaranteed outcome.

People

Eric Smith

Team Member

Grants Received– no grants recorded

Add grant

Funding Asks

grantmaking.ai Launch Round

Applied

Minimum

$20,000

Ideal

$250,000

How the money will be spent

$20,000:

My time to conduct research and coordinate the project
Interviews with stakeholders (accessibility experts, standards organizations, AI safety researchers, credential providers, and end users)
Development of an open-technical specification for the Adaptive Layer
A limited proof-of-concept demonstrating user-controlled adaptive attributes
Privacy, security, and misuse analysis
Publication of findings, documentation, and open-source materials

$250,000+:

Production-quality reference implementations
Large user studies
In-depth work with standards bodies and open-source communities
Explore integrations with existing digital credential ecosystems
Produce additional documentation and educational resources

Discussion

No comments yet. Be the first to share your thoughts.