Introducing Idios AI Profile
A living AI Profile, built around you.
A continuously learning model that brings together your personal context, learns from interactions and feedback, and helps AI systems adapt to you—on your terms.
Teach Idios
Help Idios get to know you. Start with what you already know, bring relevant context from other AI systems, or explore your preferences together.
- Start with what you know: import selected AI context, or tell Idios your goals, preferences and boundaries yourself.
- Learn with Idios: a guided conversation about what matters to you, or realistic situations that reveal your trade-offs.
- Questions worth asking: Idios asks only when an uncertainty matters, and says why. Skip anything.
- Type or speak your answers. You review every answer before it’s kept, and can correct what Idios learns.
My Living Model
Inspect, verify, correct, and control your personal understanding.
- Stable tendencies, kept apart from preferences that hold only in one domain.
- Temporary context that expires on its own.
- The evidence behind each inference, and how certain Idios is.
- Verify or correct any item, or mark it do-not-learn.
Experience Adaptation
See how your profile can shape AI responses and recommendations.
- See which permitted profile items shaped a response.
- React in a click; your feedback updates the model.
- Watch the next response change because of it.
How Idios learns
Learning that improves through experience.
Idios is designed to learn more than facts about you. It forms hypotheses from personal evidence, identifies what it still needs to learn, and refines its understanding through interactions and feedback.
-
01
Hypothesize
Discover meaningful patterns.
Interpret conversations, choices, and personal context while distinguishing what is known, inferred, or uncertain.
-
02
Ask / Test
Learn what matters next.
Ask targeted questions or explore low-stakes scenarios when additional evidence could improve understanding.
-
03
Update
Refine contextual understanding.
Update the living user model, including when preferences apply, how certain they are, and where they may generalize.
-
04
Adapt & Learn
Improve through experience.
Help AI systems adapt to the individual, then learn from relevant outcomes, corrections, and feedback.
In practice
One preference. Two different decisions.
“I prefer fewer options only when I’m making a quick decision.”
The preference is conditional, not universal.
A quick decision
Pick a lunch spot near the office.
Present a concise shortlist of two relevant options.
A considered decision
Help me choose a laptop for work.
Don’t automatically limit the options. Offer a broader comparison or clarify the desired level of detail.
Same individual. Different contexts. More appropriate adaptation.
Our product goal
The more you use Idios, the less you need to correct AI.
We’re working toward AI that requires fewer repeated explanations, unnecessary questions, and avoidable corrections as it learns about you.
This is a goal we’re designing toward and intend to evaluate—not a measured result we’re claiming today.
Your control
Your living profile. Yours to shape.
Continuous learning should come with clear boundaries. We’re designing Idios so you can understand and guide what it learns about you—and how that understanding is used.
Choose what it learns
Control which sources and personal context Idios can learn from, what it retains, and what remains off-limits.
Review and refine
See what Idios knows or infers, understand the supporting evidence, correct mistakes, and update preferences as your circumstances change.
Control how it’s used
Decide what relevant understanding can be shared with authorized AI systems, for which purposes, and under what conditions.
Continuously learning. Always yours to steer.