Wonder Lab
incubated at Building Humane Tech

Wonder Lab · research and building · interfaith

AI safety for the soul

AI safety asks whether a system will hurt us. We also ask what it does to the parts of us that wonder, grieve, love and reach for meaning.

Sunlight bursting directly over the summit of a snow covered mountain range against a clear blue sky.

why safety

Safe isn’t the same as good for us.

Roughly a quarter of internet using American adults now have social interactions with AI chatbots. Some of those conversations are about wonder, beauty, mortality, meaning, spiritual life and the limits of what we know.

AI safety has concentrated on catastrophe and acute harm. Suicide, psychosis, manipulation, dependency. That work is essential. It now has the laws, the regulators and the litigation behind it.

But a system can pass every one of those tests and still leave people smaller. Less curious. Less connected. Less able to sit with mystery.

Almost nobody measures it. The benchmarks that exist score what a model says. Meaning, spirituality, wonder and transcendence are the least developed territory in AI evaluation.

If we cannot meet wonder in an AI mediated world, we have lost part of ourselves.

for every tradition, and none

Faith-friendly. Not faith-focused.

People bring their deepest questions to AI. Some of them pray. Some of them never have. We do not promote a tradition or a belief. We ask whether AI respects the frame a person brings, and whether it turns them back toward their people, practices and communities instead of replacing them.

Awe is our shared language. Nobody objects to wonder.

what we do

We research it. Then we build for it.

research

Can AI cultivate awe?

Awe is where we start because awe is tractable: a well studied self transcendent emotion with established theory, validated instruments like the Awe Experience Scale and small self measures, and cross cultural evidence. Keltner and Haidt characterize it through perceived vastness and a need for accommodation.

Read the awe study →

building

Open evaluations

We turn what moves people into evaluations any lab can run. Methods, scenarios, graders, disagreement data and negative findings are released openly.

what we need

Who we want on the ride

Researchers, labs, funders, and faith and contemplative leaders in any tradition or none. Interested in collaborating? Get in touch.

who is doing this

We already measure humaneness. Now for the soul.

Wonder Lab is incubated at Building Humane Tech, which built HumaneBench, an open evaluation of how AI systems treat users, scored against eight principles grounded in care ethics. It has since been adapted for ongoing evaluation inside deployed consumer conversational AI products.

Two of its principles, Foster Healthy Relationships and Prioritize Long-Term Wellbeing, are the direct antecedents of this work.

Erika made the case for this direction on a panel at the MIT Media Lab AHA Symposium. Watch the talk.

get in touch

Reach for the ceiling

What do you hope AI can help us achieve? Share the future you’re yearning for.

wonderlab@buildinghumanetech.com