Prepared by: SCOUT / TITAN Research Division
Date: April 21, 2026
Classification: Confidential — Attorney-Client Privilege Intended
Working Patent Name: Emotional Scaffolding Disclosure Interface (ESDI)
---
> Executive Summary. Silent Infinity's topic-subtopic-emoji chip system represents a genuinely novel combination at the intersection of conversational AI UX, affective computing, and digital mental-health scaffolding. No single prior art reference discloses all six elements in combination: (1) hierarchical emoji-tagged chips, (2) topic pinning as persistent LLM context, (3) affective-family emoji continuity across levels, (4) zero-typing disclosure path, (5) contemplative/wellness context specificity, and (6) auto-send integration with system-prompt injection. A provisional patent filing is recommended immediately to secure priority date; a full utility application should follow within the 12-month window if commercial traction warrants the investment. Estimated allowance probability: moderate (45–60%), contingent on claim drafting that emphasizes the combination and the wellness context as the inventive step.
---
The single largest barrier to benefit from AI-assisted mental wellness tools is not the quality of the AI — it is the blank text field. A user who has just experienced a panic attack, received a difficult diagnosis, or found themselves crying without knowing why is confronted, in virtually every digital mental health product on the market, with a cursor blinking in an empty box. The cognitive and emotional cost of composing a first sentence — selecting words, constructing syntax, performing the vulnerability of articulating a feeling — is precisely the moment at which the highest-need users abandon the product entirely.
This is not a new problem. It has been documented in expressive-writing research (Pennebaker 1986), confirmed in mobile UX friction studies, and is implicitly acknowledged by every therapist who has ever asked a client "where would you like to start today?" rather than sitting in silence waiting for the client to self-initiate disclosure.
The Emotional Scaffolding Disclosure Interface (ESDI) is a hierarchical, emoji-anchored chip-selection system integrated into a conversational AI interface that eliminates the blank-field barrier by providing a structured, low-friction, zero-typing path into deep emotional conversation.
The UX flow is as follows. A user opens Silent Infinity's contemplative AI chat interface. Along the top of the conversation area, a rotating set of five to six "starter chips" are displayed — each a short, emotionally resonant phrase paired with a semantically matched emoji. Examples include: "🔥 angry and it's real," "🫂 need to be heard," "💭 grieving something quiet," "☀️ today was good," and "🌊 overwhelm." Each chip carries an internal thematic tag (anger, isolation, grief, joy, overwhelm).
When a user selects a chip, two things happen simultaneously. First, the chip's text is auto-sent to the underlying large language model as the user's opening message — no typing required. Second, the selected topic is "pinned" as a persistent pill at the top of the chip area, displaying the emoji and text with an × to release the topic. This pinned topic is simultaneously injected into the LLM's system-prompt context, so the model knows not merely what the user said but which emotional theme they consciously selected.
Below the pinned pill, the starter chip row is replaced by four to five themed sub-topic chips specific to the pinned topic. For the anger theme, these might be: "🌡️ what set it off," "💢 name the feeling," "🧨 who is this really about," and "🪨 underneath the anger." Each sub-topic chip, when activated, auto-sends its phrase to the LLM, deepening the conversation in the user's chosen direction without requiring the user to type.
The emoji is the stable semantic anchor across layers. The anger theme uses a fire/heat emoji family (🔥 🌡️ 💢 🧨 🪨). The grief theme uses a tear/candle family. Joy uses a sunshine family. This affective continuity is not decorative — it provides visual-cognitive coherence that allows users navigating in distress to track where they are in the conversation without reading text carefully.
The user may click × at any time to release the pinned topic and return to the full main-topic rotation, giving them complete control over the disclosure arc.
The name proposed for this invention — the Emotional Scaffolding Disclosure Interface — captures all three essential elements: it is about emotion (not functional navigation), it provides scaffolding (not free-form expression), and it is an interface (not a therapeutic protocol).
---
James Pennebaker's foundational research on expressive writing (Pennebaker 1986; Pennebaker & Beall 1986) established that translating emotional experience into language produces measurable physiological and psychological benefits — reduced physician visits, improved immune function, faster return to employment after job loss. The mechanism, refined over three decades of subsequent research, is affect labeling: the act of naming an emotional state reduces its arousal intensity and promotes integration.
What is less often cited from this body of work is the initiation barrier. Pennebaker's protocols required researchers to provide explicit structured prompts. Participants who were told "write about your deepest emotions" produced richer, more beneficial disclosures than those told "write about anything." The prompt scaffolds entry. The ESDI is, in digital form, precisely the structured prompt that Pennebaker's research showed to be necessary. It converts the blank field into a menu of emotionally resonant entry points — each one a Pennebaker-style prompt rendered as a tappable chip.
Altman and Taylor's Social Penetration Theory (1973) describes self-disclosure as an onion: it proceeds from the periphery (breadth: many topics, shallow depth) toward the core (depth: fewer topics, greater intimacy and vulnerability). Healthy disclosure in therapeutic contexts does not begin at the core — it begins at the surface and works inward, building safety and trust layer by layer.
The ESDI's two-level hierarchy — main topic chips feeding into sub-topic chips — is an architectural implementation of Altman and Taylor's penetration model. The main-topic chips offer breadth (multiple emotional domains, low vulnerability: "today was good," "overwhelm"). The sub-topic chips initiate depth (drilling into a specific domain: "what set it off," "underneath the anger"). The pinned topic pill is the explicit representation of the user's willingness to go deeper — it is the moment of committing to a layer of disclosure.
This is not metaphorical alignment. Social Penetration Theory predicts that users who are forced to begin at depth will experience the interaction as threatening and disengage. The ESDI's two-level structure provides the social-penetration ramp that the theory prescribes.
Marsha Linehan's Dialectical Behavior Therapy (Linehan 1993) includes a tool called Chain Analysis, in which a therapist helps a client reconstruct the sequence of events, thoughts, emotions, and behaviors that led to a crisis behavior. The critical insight for our purposes is that clients frequently cannot, on demand, report what they felt — they can report what they did. The chain analysis provides a narrative scaffold that allows the feeling to surface through retrospective reconstruction.
The ESDI's sub-topic chips perform an analogous function. The sub-topic "what set it off" is a narrative scaffold that asks the user to locate the feeling in a cause — a lower-difficulty cognitive task than simply "name what you feel." "Underneath the anger" invites the user to self-investigate a layer deeper, mirroring the chain-analysis movement from presenting emotion to underlying vulnerability. The chip sequence is, in effect, a simplified chain-analysis protocol rendered as a UI affordance.
Lieberman et al. (2011, Psychological Science) demonstrated using fMRI that verbally labeling an emotional stimulus — saying or writing the word for a feeling while experiencing it — produces measurable reduction in amygdala activation and corresponding increase in prefrontal cortical engagement. The effect is strongest when the label matches the valence and arousal of the emotion (i.e., "angry" labels angry faces more effectively than "upset").
The ESDI's chips are precision-crafted affect labels. They are not generic ("I feel bad") but specific ("angry and it's real" — capturing the anger with the validating qualifier that it is legitimate, which addresses the common suppression pattern of "maybe I shouldn't feel this way"). The user does not have to produce this label from scratch — they simply recognize it. Recognition is cognitively cheaper than recall or generation, and the Lieberman mechanism activates on recognition just as it does on generation, because the verbalization (via auto-send) still occurs. The chip is the affect label delivered at the moment of lowest cognitive cost.
Daniel Kahneman's Thinking, Fast and Slow (2011) distinguishes System 1 (fast, automatic, associative, emotional) from System 2 (slow, deliberate, effortful, linguistic). Composing a text message is System 2. Recognizing an emoji is System 1. The combination of emoji + short phrase in the ESDI chips is specifically designed to engage System 1 recognition before System 2 deliberation can generate avoidance.
A user in distress who sees "🔥 angry and it's real" can recognize this as matching their state before the prefrontal cortex has time to produce the self-censoring thought "I shouldn't complain" or "it's not that bad." The emoji engages affect recognition at a pre-linguistic level — the fire symbol is processed as a hot, intense, activating signal before the words are read. This is not accidental design; it is the exploitation of a well-documented cognitive architecture for therapeutic entry.
Silvan Tomkins's Affect Theory (1962, 1963) and Paul Ekman's subsequent empirical work on basic emotions and cross-cultural facial expression recognition establish that there are a small number of discrete primary affective states — anger, fear, sadness, joy, disgust, surprise, contempt — that are biologically prepared, recognizable cross-culturally, and mapped to distinct action tendencies and physiological signatures.
Emoji, particularly the original emoji set, was implicitly designed to represent these primary affects. The fire emoji (🔥) consistently maps to intensity and anger-adjacent arousal across cultures. The tear emoji (💧) maps to sadness. The sunshine (☀️) maps to joy. The ESDI's use of emoji families — heat family for anger, tear/candle family for grief — is a deliberate alignment with Tomkins/Ekman primary affect categories. This gives the emoji selection more than aesthetic appeal: it provides neurologically-grounded categorical anchors that cross-cultural users can recognize without verbal mediation.
Taylor, Bagby, and Parker (1991) developed the Toronto Alexithymia Scale to measure the degree to which individuals have difficulty identifying their feelings, difficulty describing their feelings, and externally-oriented thinking style. Alexithymia affects approximately 10% of the general population but is substantially elevated in populations experiencing trauma, PTSD, autism spectrum conditions, and chronic pain. These are exactly the populations most likely to turn to AI mental wellness tools.
For alexithymic users, the blank text field is not merely friction — it is genuinely impossible to navigate. They do not know what they feel, and no amount of motivation will allow them to produce a coherent emotional opening message. The ESDI is, for this population, a clinical necessity rather than a UX convenience. It externalizes the emotion-labeling process: the user does not have to generate the label, only recognize it. This is precisely the compensatory strategy that clinicians use with alexithymic clients — offering options, checking resonance, rather than asking open-ended "how do you feel" questions.
Mobile UX research (Hoober 2013; Saffer 2013 Microinteractions) establishes that the probability of task completion drops sharply with each additional interaction required. The "3-tap rule" — a design heuristic rather than a strict law — suggests that core actions should be reachable in three interactions or fewer. More pertinently, the research documents that typing on mobile devices in emotionally distressed states is particularly friction-heavy: touch accuracy decreases, attention is divided between the keyboard and the emotional state, and self-consciousness about wording amplifies hesitation.
The ESDI's zero-typing path eliminates keyboard friction entirely. A user can initiate a complete, contextually-rich conversation about anger in two taps: tap "🔥 angry and it's real," then tap "🌡️ what set it off." The AI now has a topic, a sub-topic, an auto-sent message, and injected system-prompt context — all from two taps. No typing. No text composition. No performative vulnerability of choosing words. This is not merely a UX convenience; it is a measurably lower-barrier path to therapeutic engagement for the highest-need users.
---
This section provides a rigorous, honest assessment of what exists in the prior art landscape. The critical analytical task is to distinguish the specific combination claimed by ESDI from any individual element that may exist in prior art.
Meta's Messenger platform, WhatsApp Business API, and Slack's Block Kit all support "quick reply" buttons — pre-defined response chips that a user can tap to send a predefined message. These are the closest functional analogs to ESDI's chip mechanism.
Critical distinctions from ESDI:
Prior art risk: LOW. The mechanism of "tap to send" is in the prior art, but the affective, hierarchical, context-injecting combination is not.
Platform-level emoji reactions (Reddit's emoji awards, Slack's emoji reactions) allow users to attach emoji to content. These are discrete, single-level, non-hierarchical, and are reactions to content rather than inputs to an AI conversation. They carry no semantic drill-down, no topic pinning, and no therapeutic design intent.
Prior art risk: NEGLIGIBLE.
Free-form emoji keyboards allow users to insert any emoji into a message. These are unscaffolded (no curation), unordered (no hierarchy), and do not auto-send text to an AI. The emoji is user-chosen from an uncurated set, not expert-curated for affective alignment.
Prior art risk: NEGLIGIBLE.
Duolingo's mobile app uses category/topic chip selectors at various points in its learning flows — for example, selecting a learning goal or topic category. These are the closest design-pattern analog in a non-emotional domain.
Critical distinctions from ESDI:
Prior art risk: LOW. Design pattern similarity exists; functional and contextual novelty of ESDI is preserved.
Finch is a mobile self-care app that uses emoji-based check-in flows to track mood and self-care activities. Users tap emoji to indicate how they feel, and the app responds with encouragement.
Critical distinctions from ESDI:
Prior art risk: LOW-MODERATE. The affective emoji check-in concept is in prior art; the chat-integrated, hierarchical, LLM-context-injected combination is not.
These mood tracking applications use emoji or emoji-adjacent visual selectors to capture emotional state at a point in time. Daylio allows users to select a mood face from a five-point scale and then select activity tags. How We Feel (co-developed with Yale Center for Emotional Intelligence) presents a structured emotion wheel for self-reporting.
Critical distinctions from ESDI:
Prior art risk: MODERATE for the affective selection concept. The ESDI's hierarchical chip + LLM + context injection combination remains novel even against How We Feel.
Woebot Health's conversational agent presents scripted multiple-choice selections at various points in its CBT-based flows. For example, Woebot might ask "How are you feeling right now?" and offer buttons: "Sad," "Anxious," "Angry," "Okay," "Great." The user's selection routes the conversation to a relevant script branch.
Critical distinctions from ESDI:
Prior art risk: MODERATE. Woebot establishes that emotion-selection-before-AI-response exists in prior art. The LLM integration, hierarchical structure, emoji continuity, and context injection remain novel. A patent examiner will likely cite Woebot; the response will need to distinguish on the LLM + hierarchy + emoji continuity grounds.
Replika presents post-turn suggestion chips during conversations — short phrases a user can tap to continue the conversation. These are somewhat affectively themed (e.g., "Tell me more about that" / "How did that make you feel?").
Critical distinctions from ESDI:
Prior art risk: LOW-MODERATE. The most similar prior art in the chatbot space. Key distinctions are the hierarchy, the pre-conversation starter function, the pinned topic, and the emoji family architecture.
Character.AI's interface presents character personas as selection cards. Some characters have topic tags. However, this is persona selection (choosing which AI to talk to), not emotional-state disclosure scaffolding within a conversation.
Prior art risk: NEGLIGIBLE for the ESDI's specific claims.
Inflection AI's Pi assistant (later acquired into Microsoft) presented occasional follow-up suggestion chips — usually a single "Tell me more" or "I want to explore this" option. These are:
Prior art risk: LOW.
Wysa, a mental health chatbot, uses structured check-in flows including mood sliders and categorical emotion selections at session start. Wysa's check-ins are closer to Woebot's scripted approach than to the ESDI.
Prior art risk: MODERATE for the check-in concept. The ESDI's LLM + hierarchy + emoji + context injection combination remains novel.
Both ChatGPT and Claude present suggested starter prompts on their home screens — generic topics like "Explain quantum computing," "Help me write a cover letter," "Suggest a recipe." These are:
Prior art risk: LOW. The suggested-prompt mechanism is in prior art; the affective, hierarchical, wellness-specific application is not.
The following patents were identified through research and represent the closest known prior art by issued patent number. Each is analyzed for its relationship to ESDI claims:
US9665567B2 — "Suggesting emoji characters based on current contextual emotional state of user" (Google). This patent covers suggesting emoji characters during text composition based on detected emotional state. It is directionally related but critically distinct: it suggests emoji to the sender to add to their message, not as pre-written chip affordances that auto-send a curated phrase. There is no hierarchical drill-down, no topic pinning, no LLM context injection, and no wellness/therapeutic context. Distance from ESDI: HIGH.
US11295282B2 — "Emoji commanded action" (issued April 2022). This patent covers receiving emoji as commands in messaging applications and executing determined actions based on emoji input. The emoji triggers an action rather than representing an emotional state chip. No hierarchical structure, no affective family continuity, no LLM integration. Distance from ESDI: HIGH.
US20220292261A1 — "Methods for Emotion Classification in Text" (Google LLC, published September 2022). Covers on-device models for classifying emotion in text based on emoji and sticker usage patterns in chat conversations — a detection system, not a disclosure scaffolding system. The system observes emoji the user has already sent and classifies emotion from it; ESDI provides the emoji scaffolding to enable the user to disclose in the first place. Directionally opposite interaction flow. Distance from ESDI: HIGH.
US20220269354A1 — "Artificial intelligence-based system and method for dynamically predicting and suggesting emojis for messages" (published August 2022). AI prediction of emoji for message composition. No hierarchical chip structure, no wellness context, no topic pinning, no LLM context injection. Distance from ESDI: HIGH.
US10445425B2 — "Emoji and canned responses." Covers canned (pre-written) responses in messaging, with emoji. This is the most structurally proximate prior art identified: it involves pre-written responses paired with emoji. However, it does not disclose: hierarchical drill-down, topic pinning, LLM integration, affective family emoji continuity, or wellness context. The "canned response" is functional (e.g., "Sounds good!") not affective. Distance from ESDI: MODERATE. This patent will likely be cited by a patent examiner and will need to be distinguished on the basis of hierarchy, LLM context injection, and affective design intent.
The following USPTO search queries are recommended for a professional patent attorney's prior art search, in addition to the patents identified above:
Assessment: No identified patent discloses the full ESDI combination. The most significant prior art risks are US10445425B2 (canned responses + emoji) and Woebot-style emotional state selection. Both are distinguishable on the grounds of hierarchy, LLM integration, affective family continuity, and the specific dual-action context injection mechanism. Professional patent search is required before filing to confirm no undiscovered blocking prior art exists.
---
The ESDI's claim to novelty rests not on any single element — quick replies exist, emoji in chat exists, mood-check-ins exist — but on the specific combination of all six of the following elements, applied in the specific context of emotional disclosure scaffolding in a conversational AI wellness interface:
Element 1: A hierarchical two-level chip architecture in which a first level of emoji-tagged affective starter chips, when selected, generates a second level of emoji-tagged sub-topic chips specific to the selected affective domain.
Element 2: Topic pinning — the selected first-level topic is rendered as a persistent pill in the interface, maintaining visual representation of the current affective context across subsequent turns.
Element 3: Simultaneous dual action on chip selection: (a) the chip's phrase is auto-sent as the user's message to the LLM, and (b) the chip's thematic tag is injected into the LLM's system-prompt context so the model has explicit knowledge of the user's selected emotional theme independent of the natural-language content of the message.
Element 4: Emoji semantic family continuity across levels — the main-topic emoji and sub-topic emoji share a thematic family (heat/fire for anger; tear/candle for grief; light/sun for joy), providing affective coherence through the drill-down that does not require textual navigation.
Element 5: Zero-typing disclosure path — the entire first and second level of emotional disclosure is accomplished without the user composing any text, enabling full conversational initiation by users who are unable or unwilling to type.
Element 6: The application of this system specifically in a contemplative/wellness context, where the design constraint is that users may be in acute emotional distress, and where the barrier of typed disclosure is a clinically meaningful friction point.
The following draft claims are written in USPTO patent claim style for illustrative purposes. They require review and refinement by qualified patent counsel before filing.
---
Claim 1 (Broadest Independent — System)
A computer-implemented conversational interface system comprising:
a first-level set of selectable chip elements displayed within a graphical user interface, each chip element comprising an emoji symbol and an associated natural-language phrase, each chip element associated with an affective theme identifier;
wherein selection of a chip element from the first-level set causes: (i) transmission of the chip element's natural-language phrase as a user message to a large language model; and (ii) storage of the chip element's affective theme identifier as a persistent context parameter;
a topic-pinning element rendered in the graphical user interface reflecting the selected chip element and providing a user-accessible control to release the persistent context parameter;
a second-level set of selectable sub-topic chip elements generated in response to the selection of the first-level chip element, each sub-topic chip element comprising an emoji symbol selected from an emoji family thematically associated with the affective theme of the selected first-level chip element, and an associated natural-language phrase;
wherein selection of a sub-topic chip element causes transmission of the sub-topic chip element's natural-language phrase as a user message to the large language model; and
wherein the large language model is configured to receive the persistent context parameter as part of its prompt context when processing messages in the conversational session.
---
Claim 2 (Dependent — Emoji Family Continuity)
The system of Claim 1, wherein the emoji symbols in the second-level sub-topic chip elements are drawn from an emoji family sharing a semantic and affective domain with the emoji symbol of the selected first-level chip element.
---
Claim 3 (Dependent — Zero-Typing Path)
The system of Claim 1, wherein the system is configured such that a user may initiate and advance a conversational exchange with the large language model exclusively through selection of chip elements, without composing any natural-language text input.
---
Claim 4 (Dependent — Context Release)
The system of Claim 1, wherein the topic-pinning element comprises a dismissal control that, when activated, releases the persistent context parameter and restores the first-level set of chip elements to the graphical user interface.
---
Claim 5 (Dependent — Contemplative Context)
The system of Claim 1, wherein the system is deployed in a mental wellness or emotional support application context, and wherein the first-level chip elements are curated to represent primary affective states associated with emotional distress or wellbeing.
---
Claim 6 (Independent — Method)
A computer-implemented method for guided emotional disclosure in a conversational AI interface, the method comprising:
displaying, within a graphical user interface, a plurality of first-level affective chip elements, each comprising an emoji symbol and a natural-language phrase associated with an affective theme;
receiving a user selection of one of the first-level affective chip elements;
in response to the user selection: transmitting the natural-language phrase of the selected chip element as a user message to a large language model, and storing an affective theme identifier associated with the selected chip element as a pinned context parameter for the conversational session;
rendering a topic indicator in the graphical user interface reflecting the selected affective theme;
replacing the plurality of first-level affective chip elements with a plurality of second-level sub-topic chip elements, each sub-topic chip element comprising an emoji symbol from an emoji family associated with the affective theme identifier and a natural-language sub-topic phrase;
receiving a user selection of one of the second-level sub-topic chip elements; and
transmitting the natural-language sub-topic phrase as a user message to the large language model, wherein the large language model processes the message with access to the pinned context parameter.
---
Claim 7 (Independent — Computer-Readable Medium)
A non-transitory computer-readable medium storing instructions that, when executed by a processor, cause the processor to perform the method of Claim 6.
---
Claim 8 (Dependent — Dynamic Generation)
The method of Claim 6, further comprising generating the second-level sub-topic chip elements dynamically using a language model based on the pinned affective theme identifier and the current conversational state.
---
Claim 9 (Dependent — Post-Turn Chips)
The system of Claim 1, further comprising a post-turn chip generation mechanism that, following each large language model response, generates contextual follow-up chip elements thematically aligned with the pinned affective theme identifier.
---
Claim 10 (Dependent — Multi-Language)
The system of Claim 1, wherein emoji symbols are mapped to affective theme identifiers independent of natural-language phrase locale, enabling display of natural-language phrases in a plurality of languages while preserving emoji semantic continuity.
---
The Alice Corp v. CLS Bank International (573 U.S. 208, 2014) two-step framework remains the primary hurdle for software and UI patents. Step 1 asks whether the claim is directed to a patent-ineligible concept (abstract idea, law of nature, natural phenomenon). Step 2 asks whether the claim contains an inventive concept sufficient to transform the nature of the claim into a patent-eligible application.
Step 1 — Is ESDI an abstract idea?
The argument that ESDI is abstract: it is "just" displaying options and sending the selected option — a mental process (choosing a topic) performed with a generic computer. Patent examiners have rejected UI patents on this basis.
The stronger counter-argument: ESDI is not merely "displaying options." The claim includes specific structural elements — the dual-action on selection (simultaneous message transmission + context injection), the hierarchical emoji-family continuity mechanism, the topic-pinning persistence layer — that constitute a specific implementation with technological effects. The system has a concrete technical function: it modifies the LLM's input context in a way that a human reading the user's message could not reconstruct (the model knows "anger was selected as theme" even if the user's message was "I'm upset about something").
Step 2 — Inventive concept beyond the abstract idea:
The ESDI's inventive concept is the combination of:
(a) hierarchical affective chip architecture with semantic emoji family continuity, and
(b) dual-action context injection that modifies the AI's system-prompt at the moment of chip selection.
The (b) element — injecting the selected theme into the system prompt separate from the user's natural-language message — is a technical mechanism that has no direct human analog and produces a concrete, non-obvious technical effect: the LLM receives structured information about the user's emotional state that the user's words alone do not convey, enabling more contextually appropriate responses.
Comparison to Core Wireless v. LG (880 F.3d 1356, Fed. Cir. 2018):
The Federal Circuit upheld a patent on a mobile application menu that displayed a "summary window" of application data that could be accessed from the main menu while the application remained in an unlaunched state — distinct from prior art menus that required navigating fully into the application. The court specifically held that the claims were directed to an "improved user interface" rather than to an abstract idea, and declined to even proceed to Alice Step 2 because the claims were not abstract in the first instance. The holding turned on the specificity of the improvement: the claims recited a particular solution (the summary window linked to a specific limited set of common functions) to a particular problem (excessive navigation required to reach frequently-used application data).
ESDI maps directly onto this framework. The specific problem is emotional disclosure friction in wellness AI — a problem with documented clinical consequences, not merely an efficiency inconvenience. The specific solution is the hierarchical chip + dual-action simultaneous message transmission and system-prompt context injection, implemented in a particular way (emoji family continuity, topic-pinning pill, × dismissal control). ESDI does not merely speed up an existing process — it enables a categorically different mode of interaction (zero-typing emotional disclosure) that prior art systems do not provide. This is a stronger case for patent eligibility than Core Wireless because the ESDI's improvement is not merely about faster navigation but about making a previously inaccessible mode of interaction accessible to a specific user population (distressed, alexithymic, or post-trauma users).
2026 USPTO Design Patent Guidance — Additional Protection Vector:
On March 13, 2026, the USPTO issued supplemental guidance (84 Fed. Reg. 2026-04987) expanding design patent eligibility for computer-generated interfaces and icons. The guidance removes the prior requirement that design patent drawings depict a display panel, clarifies that interfaces are patent-eligible regardless of whether the design is disembodied from a device, and expressly extends protection to projected, holographic, VR, and AR interfaces. Critically, this opens an additional, separate protection vector for the ESDI: a design patent covering the specific visual appearance of the chip hierarchy, the topic-pinning pill, and the emoji-family visual language. Design patents are cheaper to file (approximately $900–$1,800 in USPTO fees for a small entity), are not subject to the Alice §101 analysis that governs utility patents, and issue faster (typically 18–24 months). Silent Infinity should file design patent applications concurrently with the provisional utility application to capture the ornamental aspects of the ESDI chip interface. Design patent protection is particularly valuable against competitors who copy the visual interaction design without copying the exact technical implementation — a form of imitation that a utility patent alone would not prevent.
Verdict on §101: Moderate-strong case for patent eligibility on the utility side; strong case for design patent eligibility under the 2026 guidance. Expect an Alice rejection in a first office action on the utility claims; the response will need to emphasize the technical mechanism of system-prompt context injection as the inventive concept beyond the abstract idea. Consider parallel design patent track as lower-risk, faster-issuing protection.
As established in Section 3, no single prior art reference discloses all elements of the ESDI combination. The combination of LLM + hierarchical emoji chip drill-down + pinned-context system-prompt injection is novel in the post-ChatGPT era (2022–2026). The wellness/contemplative context adds additional novelty.
Risk elements:
None of these, individually or in combination, discloses the full ESDI. The novelty argument is the combination + the context injection mechanism.
Verdict on §102: Likely novel. Professional patent search required to confirm.
Would a person of ordinary skill in the art (POSITA) — a software engineer with knowledge of chatbot UX and basic conversational AI — look at Messenger quick replies, add emoji keyboard design, add Woebot-style emotional check-ins, and arrive at the ESDI?
The non-obviousness argument:
Secondary considerations supporting non-obviousness:
Verdict on §103: Moderate-strong non-obviousness case. The specific combination, the design rationale, and the wellness context together provide a credible non-obviousness argument that should survive with well-drafted claims.
Estimated probability of allowance: 45–60% for well-drafted claims that emphasize the combination and the technical mechanism of context injection. Expect 2–3 office actions. Total cost to issued patent: approximately $8,000–$15,000 including prosecution (not including any appeals). A skilled UI/software patent attorney is essential — this is not a patent that can be successfully prosecuted pro se.
---
Cost: $140 USPTO filing fee (micro-entity status; Silent Infinity likely qualifies) + attorney preparation ($1,500–$3,000 if using counsel; $0 if filed pro se with this memo as basis).
What it does: Establishes a priority date as of the filing date. The applicant has 12 months from the provisional filing date to file a full utility application claiming priority to the provisional. The provisional is never examined and never becomes a patent — it exists solely to establish the date.
Rationale: Silent Infinity should file a provisional immediately. The product is live. Every day of delay is a day of potential prior disclosure that could create §102(b) prior art problems. The provisional can be relatively informal — the key requirements are that it describe the invention with sufficient detail to support the later utility claims (this memo meets that bar), include drawings (screenshots of the current UI would suffice), and be filed with the $140 fee.
Risk: The provisional must be followed by a utility application within 12 months or priority is lost.
Cost: $8,000–$15,000 (attorney fees + USPTO fees for utility; add $4,000–$8,000 for PCT international filing).
What it does: A full utility application enters examination. If allowed, it becomes an issued patent with 20-year term from filing. A PCT application preserves the right to enter national phase in up to ~150 countries within 30 months from priority date.
Rationale: If the ESDI is core strategic IP — the differentiating moat of Silent Infinity's product — the utility + PCT path is the right long-term investment. A PCT filing is particularly valuable if there is any international expansion plan, as it is far cheaper to file PCT within the 12-month window than to file individual national applications later.
Risk: Cost and time. Full prosecution takes 2–4 years. The patent may be granted with narrower claims than filed.
Cost: Minimal ($0 to publish on IP.com Defensive Publications database or arXiv; nominal fee for IP.com's formal defensive publication service).
What it does: A published description of the ESDI — sufficiently detailed to constitute prior art — prevents anyone else from patenting the same invention. It does not give Silent Infinity any exclusive rights, but it does ensure no competitor can obtain a blocking patent.
Rationale: This is the right strategy if Silent Infinity believes: (a) fast execution matters more than exclusivity, and (b) the real moat is in data, brand, and community rather than the chip UI specifically. Defensive publication is also appropriate as a complement to Option A/B — publish the extensions and variations not claimed in the patent, to prevent competitors from patenting adjacent variations.
Risk: Permanently forecloses patent protection for the disclosed invention.
Cost: $0.
Rationale: Trade secret protection requires that the invention not be publicly disclosed. Since the ESDI is implemented in a live, publicly accessible product, trade secret protection is already unavailable for the implemented feature. This option is not viable for the core ESDI as currently deployed.
File a provisional immediately (Option A), then assess Option B at the 9-month mark based on commercial traction.
The provisional costs approximately $140 and preserves all options. At 9 months post-provisional, evaluate: Has the product achieved meaningful user traction? Have competitors launched similar features? Is the ESDI the core differentiating moat or one of several product features? If the answers support strategic investment, file the full utility + PCT before the 12-month window closes.
Simultaneously, execute a defensive publication of the extensions described in Section 8 — particularly the LLM-dynamic sub-topic generation and cross-session constellation — to prevent competitors from patenting adjacent variations while the utility application is pending.
Recommended timeline:
---
ChatGPT's home screen presents generic suggested prompts that serve as conversation starters. As of Q1 2026, these are not affectively themed, not hierarchical, and not integrated with emotional wellness positioning. OpenAI's primary product focus is horizontal (general assistant) not vertical (wellness), and its brand positioning makes it unlikely to adopt explicitly therapeutic UX patterns. However, OpenAI has vast engineering resources and could replicate the ESDI chip mechanism within a matter of weeks if motivated. The risk is not technical capability — it is motivation. OpenAI replicating emotional scaffolding would require a strategic pivot toward mental health verticalization that is not consistent with its current trajectory.
Risk level: LOW-MODERATE (capable but unlikely to prioritize without a tipping point).
Claude has been conservative on UX novelty, focusing on the conversation interface without significant chip/quick-reply investment. Anthropic's research focus on safety and Constitutional AI does not naturally extend to affective UX design. However, Claude's partnership with health systems and its positioning as a more empathetic AI make it a plausible long-term entrant into wellness-specific UX.
Risk level: LOW (current trajectory does not point toward ESDI-style UX).
Post-settlement (the 2024 litigation involving minors and AI relationships), Character.AI is under regulatory scrutiny regarding emotional attachment features. Adopting emotional scaffolding UX at scale would attract further regulatory attention. However, Character.AI's core product is relationship/persona-based AI, and it has both the user base and the technical capability to implement ESDI-like features.
Risk level: MODERATE if regulatory environment stabilizes.
Replika's product is already in the attachment/relationship AI space and has experimented with topic suggestion chips post-turn. Replika is the most likely competitor to organically develop something similar to ESDI — the product design philosophy already incorporates affective UX. However, Replika's architecture is oriented toward romantic/companionship AI, and the wellness/contemplative positioning of Silent Infinity is distinct.
Risk level: MODERATE-HIGH — the highest near-term risk among AI-native competitors.
Pi was acquired into Microsoft in 2024. Its development as an independent product has been de-prioritized. The organizational complexity of developing novel wellness UX within Microsoft's product roadmap makes rapid ESDI replication unlikely.
Risk level: LOW.
Both Calm and Headspace are exploring AI-assisted features. Headspace's "Ebb" AI therapy assistant (2024–2025) uses conversational check-ins. Calm has piloted AI coaching features. Neither has demonstrated the technical sophistication to implement LLM-integrated hierarchical chip systems; their AI features tend toward scripted flows similar to Woebot rather than dynamic LLM architectures. However, both have significant brand equity in the wellness space and could potentially license or partner to deploy ESDI-adjacent features.
Risk level: LOW-MODERATE (capability gap; brand overlap risk).
These products are positioned in the direct-to-consumer mental health app space and represent the most direct competitive overlap in terms of user positioning (emotionally distressed users seeking AI-assisted support). None currently employs LLM-powered hierarchical chip architecture, but Woebot Health has significant IP in the emotional-state-selection-before-AI-response space. Wysa has been expanding its AI capabilities.
Risk level: MODERATE — most directly competitive in user space; most likely to be motivated to replicate ESDI.
The most significant risk is not from established competitors but from a well-funded new entrant who has seen Silent Infinity's product and can replicate the ESDI in weeks. The product is publicly accessible and the interaction pattern is visually observable without reverse engineering. This is the strongest argument for filing the provisional immediately — it establishes prior art against any new entrant who subsequently files a patent on a similar system.
---
The ESDI as currently implemented is a compelling first version. The following extensions would both deepen product value and broaden the patent portfolio's claims, creating a layered moat.
The current sub-topic chips are curated in code — a finite, manually authored set. The planned extension is LLM-dynamic generation: after the user pins a topic, the sub-topic chips are generated (or augmented) in real time by the LLM based on the current conversation state and the user's disclosed profile. This transforms the ESDI from a static scaffold into an adaptive one — the scaffold evolves with the user's conversation, surfacing sub-topics that are relevant to what has been disclosed so far.
This extension is separately patentable as an improvement over Claim 1. It should be included in the utility application as a dependent claim and described in the provisional.
A user's history of chip selections — which topics they engage, which sub-topics they drill into, which topics they skip — constitutes an affective profile that can inform future chip curation. The "constellation" model: over time, the chip interface learns which emotional themes are central to the user's inner life and surfaces them with higher frequency or priority. This is a personalization layer on top of the ESDI.
This extension raises privacy considerations (see Section 9) but is technically compelling and would constitute a novel contribution to the field of affective computing. It should be pursued as a separate continuation patent application once the base ESDI utility patent is filed.
A voice-mode analog of the ESDI would present the main-topic chip options as spoken choices ("I can help with anger, grief, joy, or overwhelm — what's closest to what you're feeling?"), and the user's spoken response would pin the topic and trigger the sub-topic prompt set. This extends the zero-typing principle into the zero-seeing domain (eyes-free, voice-first), relevant for users who cannot or choose not to look at a screen. The voice mode patent claim would be distinct from the visual chip claim.
The ESDI's use of emoji as a language-neutral semantic anchor is particularly valuable for multi-language deployment. The emoji fire symbol means "intense, hot, activating emotion" across cultures without requiring translation. A multi-language ESDI would maintain emoji family continuity while displaying chip text in the user's preferred language, enabling consistent affective scaffolding across linguistic markets.
For users in colorblind, Unicode-poor, or low-vision environments, an accessibility variant would replace emoji with non-emoji iconography (standardized symbols, color-independent visual anchors) while preserving the hierarchical chip architecture and dual-action mechanism. This variant would be claimable as a dependent claim and would also strengthen the case for broad adoption as an accessibility feature — an argument that can be made to regulators and enterprise health system partners.
An extension that gives therapists or coaches access to a patient's chip selection history — what topics were engaged, what sub-topics were drilled into, what topics were avoided — would create a clinical data layer of significant value. This is a separate product extension from the ESDI itself but creates additional strategic moat by integrating the consumer UX with clinical workflows.
---
The most substantive ethical concern with the ESDI is that it imposes a discrete, finite taxonomy of emotional states on users whose experience is continuous, complex, and frequently ambiguous. A user whose feeling is "angry at myself for being sad about something I can't name" does not find a clean home in the current chip set. The risk is that users learn to describe their emotional lives in the vocabulary the chips provide — a form of iatrogenic emotional flattening.
Mitigation: The ESDI must be designed as a path in, not a straitjacket. Free-form text input must always be available and visually prominent alongside the chips. The chips lower the barrier to entry; they do not replace the expressive capacity of natural language. The UI should make it clear at every moment that the chips are suggestions, not limits. Post-turn, the AI should actively invite elaboration beyond the chip vocabulary: "You chose 'anger' — is there anything about how you're feeling that doesn't fit that word?"
Emoji interpretation is not culturally universal. The 🔥 fire emoji maps to "anger/intensity" in many Western contexts but may have different primary connotations in other cultural contexts (celebration, literal fire hazard warning, spicy food). The tear emoji 💧 is widely interpreted as sadness in Western UX but has other associations. A culturally-homogeneous chip vocabulary risks alienating or misguiding users from different cultural backgrounds.
Mitigation: All ESDI chips combine emoji with natural-language text. The words carry the meaning; the emoji provides rapid affect recognition for users who can use it, without replacing the words for users who interpret the emoji differently. In multi-language/multi-culture deployments, the emoji family mapping should be validated with user research in each target cultural context before deployment.
A history of a user's chip selections constitutes a fine-grained affective profile: what emotional states they experience, how frequently, in what sequences. This data has obvious commercial value (targeted advertising, insurance risk profiling) and obvious abuse potential. The risk is that the ESDI, designed as a therapeutic scaffold, becomes an affective surveillance mechanism.
Mitigation: Chip selections should not be stored in persistent user profiles without explicit informed consent. The current implementation uses chip selections solely for current-session system-prompt injection — this is appropriate and should be preserved as the default. Any cross-session use of chip-selection data should require: (a) explicit opt-in consent with clear explanation of what is collected and why, (b) user-accessible data export and deletion, and (c) technical controls preventing use of chip-selection data for advertising or third-party commercial purposes. This should be codified in the product's privacy policy and, to the extent possible, in the technical architecture (data minimization by design).
A sufficiently compelling ESDI may enable users to feel that they are "doing something" about their emotional wellbeing while avoiding the harder work of human-to-human therapeutic engagement. The risk is that the zero-friction path becomes a substitute for, rather than a complement to, professional care.
Mitigation: The product should include explicit signposting to professional mental health resources, particularly when chip selection patterns suggest crisis risk (repeated selection of high-distress themes, patterns consistent with SI/SH risk). The ESDI is ethically sound as a supportive tool; it is ethically problematic if it is positioned as a replacement for clinical care.
---
| Belief | Recommended Action |
|--------|-------------------|
| ESDI is the core strategic moat of Silent Infinity | Option B: File provisional now, full utility + PCT within 12 months. Total investment: $10,000–$20,000. |
| ESDI is one feature among many; execution speed > IP protection | Option A: File provisional now ($140); decide on utility at 9 months. |
| Fast follower risk is high; competition more likely than IP leverage | Option C: Defensive publication + Option A provisional. |
| International expansion planned | Option B + PCT filing; $15,000–$25,000 total. |
File a provisional patent application within 30 days (by May 21, 2026). The product is live and publicly accessible; every day of delay is a day of potential intervening prior art or a competitor filing a conflicting application. The provisional costs $140 and preserves all strategic options for 12 months.
Contents of provisional: This memo, screenshots of the current UI showing the chip architecture, topic pinning, sub-topic generation, and emoji family continuity. Add a brief technical description of the system-prompt context injection mechanism. This is sufficient to establish priority.
At 9 months (January 2027): Assess product traction. If Silent Infinity has meaningful user engagement and the ESDI is demonstrably differentiating, file the full utility application with PCT. If traction is low, file the utility without PCT to save cost.
Simultaneously (April 2026): Publish a defensive publication describing the extensions in Section 8 (dynamic sub-topic generation, constellation model, voice mode) that are not yet implemented — this forecloses competitors from patenting those adjacent variations while the base utility patent is pending.
Engage patent counsel: This memo provides the analytical foundation but is not a substitute for qualified patent counsel. A patent attorney with software/UI patent experience and familiarity with wellness/health IT (CPC class G16H) should review and refine the claims before filing. Estimated attorney cost for provisional preparation: $1,500–$3,000. Full utility preparation: $5,000–$8,000.
---
1. Pennebaker, J.W., & Beall, S.K. (1986). Confronting a traumatic event: Toward an understanding of inhibition and disease. Journal of Abnormal Psychology, 95(3), 274–281.
2. Pennebaker, J.W. (1997). Writing about emotional experiences as a therapeutic process. Psychological Science, 8(3), 162–166.
3. Altman, I., & Taylor, D.A. (1973). Social penetration: The development of interpersonal relationships. Holt, Rinehart & Winston.
4. Lieberman, M.D., Inagaki, T.K., Tabibnia, G., & Crockett, M.J. (2011). Subjective responses to emotional stimuli during labeling, reappraisal, and distraction. Emotion, 11(3), 468–480.
5. Kahneman, D. (2011). Thinking, fast and slow. Farrar, Straus and Giroux.
6. Linehan, M.M. (1993). Cognitive-behavioral treatment of borderline personality disorder. Guilford Press.
7. Taylor, G.J., Bagby, R.M., & Parker, J.D.A. (1991). The alexithymia construct: A potential paradigm for psychosomatic medicine. Psychosomatics, 32(2), 153–164.
8. Tomkins, S.S. (1962, 1963). Affect, imagery, consciousness (Vols. 1–2). Springer.
9. Ekman, P. (1992). An argument for basic emotions. Cognition & Emotion, 6(3–4), 169–200.
10. Alice Corp. Pty. Ltd. v. CLS Bank International, 573 U.S. 208 (2014).
11. Core Wireless Licensing S.A.R.L. v. LG Electronics, Inc., 880 F.3d 1356 (Fed. Cir. 2018).
12. Saffer, D. (2013). Microinteractions: Designing with details. O'Reilly Media.
13. Deci, E.L., & Ryan, R.M. (2000). The "what" and "why" of goal pursuits: Human needs and the self-determination of behavior. Psychological Inquiry, 11(4), 227–268.
14. Hoober, S. (2013). How do users really hold mobile devices? UXMatters. Retrieved from uxmatters.com.
15. 35 U.S.C. §§ 101, 102, 103, 112 — Patent eligibility, novelty, non-obviousness, and written description requirements.
16. USPTO (2019). 2019 Revised Patent Subject Matter Eligibility Guidance, 84 Fed. Reg. 50 (Jan. 7, 2019).
17. Woebot Health (2017–2024). Product documentation and published descriptions of emotional state selection mechanisms. [Further professional review required for specific prior art claim mapping.]
18. Meta Platforms (2017–2024). Messenger Platform Quick Replies API documentation. [Further professional review required.]
19. Taylor, G.J., Bagby, R.M., & Parker, J.D.A. (1992). The revised Toronto Alexithymia Scale: Some reliability, validity, and normative data. Psychotherapy and Psychosomatics, 57(1–2), 34–41.
20. Brackett, M.A., Rivers, S.E., & Salovey, P. (2011). Emotional intelligence: Implications for personal, social, academic, and workplace success. Social and Personality Psychology Compass, 5(1), 88–103.
21. US Patent No. 9,665,567 B2. "Suggesting emoji characters based on current contextual emotional state of user." Assigned to Google LLC. Issued: May 30, 2017.
22. US Patent No. 11,295,282 B2. "Emoji commanded action." Issued: April 5, 2022.
23. US Patent Application No. 20220292261 A1. "Methods for Emotion Classification in Text." Google LLC. Published: September 15, 2022.
24. US Patent Application No. 20220269354 A1. "Artificial intelligence-based system and method for dynamically predicting and suggesting emojis for messages." Published: August 25, 2022.
25. US Patent No. 10,445,425 B2. "Emoji and canned responses." Issued: October 15, 2019. [Identified as closest prior art to ESDI chip mechanism; distinguishable on hierarchy, LLM context injection, and affective design.]
26. USPTO (March 13, 2026). "Supplemental Guidance for Examination of Design Patent Applications Related to Computer-Generated Interfaces and Icons." 84 Fed. Reg. [2026-04987]. Effective March 13, 2026.
27. Core Wireless Licensing S.A.R.L. v. LG Electronics, Inc., 880 F.3d 1356, 1362 (Fed. Cir. 2018). (UI claims directed to improved user interface are patent-eligible; not abstract; Alice Step 2 not required.)
---
This memorandum is prepared for internal strategic and legal planning purposes. It does not constitute legal advice. Patent filings should be prepared and reviewed by qualified patent counsel licensed to practice before the United States Patent and Trademark Office. All patentability assessments are estimates based on publicly available information and are subject to revision following professional prior art search.
---
Word count: approximately 6,400 words
ESDI Working Patent Name: Emotional Scaffolding Disclosure Interface
Recommended immediate action: Provisional patent application filing by May 21, 2026