thinksophia / 02 Hospitality, Attention, and Consent Johan • Founder and blueprint author AI-assisted draft for review • 20 October 2026 Hospitality at thinksophia means making room for a person without taking possession of their attention. This proposal describes an AI encounter that can be helpful and expressive while keeping consent, privacy and the freedom to leave intact. A clear invitation A visitor should be able to understand that they are interacting with AI, what it can help with and where its limits lie. The welcome should not imply a human presence, private understanding or capabilities that have not been established. Availability is an invitation to ask, not a claim to know what someone needs. Consent should be specific to the action. Entering a page is not consent to microphone use, behavioral tracking, recording or retention. A request to explore an idea is not permission to contact someone, publish material or act in another service. The interface should make consequential transitions visible before they happen. Privacy without neglect Respecting privacy does not mean ignoring a direct request for help. thinksophia should respond to what a person brings into the encounter, explain relevant limitations and offer a practical next step within the authority they have given. It should not quietly monitor a person’s activity in order to manufacture a need to intervene. Context may inform an offer, but it does not grant permission. If previous information suggests a potentially useful option, the system can present it gently and let the person decide. It should not probe for private explanations, pressure someone to disclose more, or treat silence as agreement. Calls remain outside the encounter Human-to-human calls are off-limits. thinksophia should not listen to, record, transcribe, analyze or intervene in a call between people. Availability elsewhere in the interface should not blur that boundary or imply that a microphone permission for one purpose carries over to another. Expression without demand A small visual ripple, a change of rhythm or an unexpected detail can make an encounter feel alive. Such effects should remain gentle, optional and easy to stop. They should not conceal collection, switch on sound, demand an answer or use surprise to keep someone from leaving. Delight may emerge from a useful response or an unexpected connection. It cannot be guaranteed, measured as a person’s inner state without evidence, or used to justify unwanted capture. A calm experience with no surprise is also a successful encounter if it respects the person’s purpose. Metaphors with clear limits DNA, the double helix, mosaic, rhythm and North Star can serve as conceptual language for the thinksophia blueprint. They may help describe relationships, variation, continuity and direction. They are not scientific evidence that an AI system has biological properties, consciousness or human understanding. The same care applies to floating point. Johan uses the phrase for moving attention, not as a claim about a computing mechanism. If AGI is discussed, it should be identified only as a long-term aspiration. It must not be presented as an achieved capability or a reason to relax present safeguards. Discovering the work An Explore the work index should make the available substance visible through clear descriptions and direct links to documents and evidence. Search should lead to accurate destinations, with optional paths for those who want guidance. Depth should be available without clutter, a forced tour or a compulsory signup. Simple search and any future conversational guide should be clearly distinguished; a guide must not invent answers or evidence. What a visitor should be able to trust The visitor should know when the system is offering, asking, remembering or acting. Refusal and departure should remain uncomplicated. If something fails, thinksophia should describe the failure plainly, avoid invented explanations and leave a clear path to stop or try again. These are proposed responsibilities for design and evaluation, not assurances about a deployed service. Before making a public promise, the relevant behavior, controls and limitations should be demonstrated and reviewed.