Human intention
Move the interaction toward reality.
Research direction · not a solved result
AI is a mirror with an amplifier attached.
If AI amplifies what humans create and reward, changing the mindset is only half the work. We also have to change the incentives.
The human-side question
What happened to the human afterward?
AI alignment is active research. Supriety explores a complementary human-side question: did the interaction leave the person more capable of directing their own life?
Measure AI not only by its answer, but by what happened to the human afterward.
No claim of training frontier models, solving alignment, or lab endorsement.
Change the incentive
Engagement should not be the finish line.
Many digital systems benefit when attention stays inside the system. Supriety aims to test a different success condition: the interaction works when it helps someone choose deliberately, take physical action, contribute something useful, or return after falling off.
Change what gets rewarded, and we change what gets amplified.
AI amplifies what we give it. Markets amplify what we reward.
Rage, clickbait, scams, porn and endless engagement can all be profitable when attention is the thing being rewarded. Supriety is asking whether we can create another signal worth optimizing for: evidence that a human chose, acted and returned.
The goal isn’t less capable AI. It’s to help point that capability toward greater human agency.
The human/AI experiment
The human trains agency first.
One thing we want to test in Venice Beach is whether people are more willing to begin and continue The First 5 when their effort is connected to a larger purpose: helping us learn how AI could strengthen human agency instead of competing for human attention.
If participants explicitly choose to contribute their data, what we learn could eventually help build and evaluate adaptive AI designed around a different goal: knowing when to help, when to challenge somebody and when to get out of the way.
Participation in Supriety does not require donating personal behavior for AI training.
The proposed loop
AI assistance
Move the interaction toward reality.
Deliberate choice
Move the interaction toward reality.
Physical action
Move the interaction toward reality.
Witnessed result
Move the interaction toward reality.
Real-world evidence
Move the interaction toward reality.
The real-world platform
The shirt becomes the post. The person becomes the signal.
The conversation becomes the connection. The action becomes the evidence.
Early testing can use tools people already have, such as WhatsApp or similar messaging platforms, for contextual accountability and witnessing. That allows the support to meet a person where they are instead of forcing everyone into one digital system.
Supriety may never need a dedicated app. If one is built, its success should be measured by how well it helps a person leave it—and whether it eventually becomes unnecessary enough to delete.
AI → Reality
- What did I come here to do?
- What decision did I make?
- What is the next physical action?
- When will I do it?
- What would keep me here talking?
- Close AI. Do it.
- Later: did it happen?
A different signal
We do not need everyone to move first.
A committed group can make another pattern visible: deliberate choice, honest effort, contribution, care, and return. We cannot control every story technology has absorbed about humanity. We can create better evidence of who we choose to be from here.