Granted Patent B2

US 12,724,978 · App. 19/540,802 · Granted Sep 1, 2026

Inventor: Scott Lipskin (Boca Raton, FL)

Assignee: The Mirror Project LLC

G06F40/30G06F40/40

Abstract

Claims (34)

1 . A computer-implemented method comprising:

receiving a first textual statement from a user, the first textual statement comprising a first human intent;

translating the first human intent into at least one of a number of optimized AI instructions and a number of trigger structures;

employing an interpretation generation engine with the at least one of the number of optimized AI instructions and the number of trigger structures in order to generate a first plurality of semantic interpretations of the first textual statement and a first corresponding number of machine-optimized prompt representations generated for each of the first plurality of semantic interpretations, and without exposing the first corresponding number of machine-optimized prompt representations to the user;

receiving a selection of a first semantic interpretation of the first plurality of semantic interpretations from the user, the first semantic interpretation comprising a clarified human intent;

generating an artificial intelligence (AI) prompt based at least in part on the selection such that the AI prompt is configured to elicit an accurate response from an AI model that is substantially devoid of hallucinations, the response being based on the clarified human intent;

sending the AI prompt to the AI model in order to generate the response;

saving the AI prompt in a storage in order to reduce repeated clarification cost associated with the interpretation generation engine and let the AI prompt compound over time;

saving the first semantic interpretation in the storage after the selection from the user;

receiving a second textual statement from the user, the second textual statement comprising a second human intent; and

remembering and utilizing the selection of the first semantic interpretation with the interpretation generation engine to generate a second plurality of semantic interpretations of the second textual statement, the second plurality of semantic interpretations comprising the first semantic interpretation because the second textual statement has a request signature matching a request signature of the first textual statement, thereby allowing the first semantic interpretation to compound over time and wherein the method further comprises logging data corresponding to the selection, whether clarification on the first plurality of semantic interpretations was requested by the user, and whether the user rated the response as helpful, and refining the interpretation generation engine over time based on the data.

2 . The method according to claim 1 , further comprising, before receiving the first textual statement from the user, receiving a selective activation of the interpretation generation engine from the user.

3 . The method according to claim 1 , before employing the interpretation generation engine:

determining whether the first textual statement has a semantic clarity level above a predetermined threshold, and either:

sending the first textual statement directly to the AI model without first employing the interpretation generation engine if the semantic clarity level is above the predetermined threshold, or

employing the interpretation generation engine to generate the first plurality of semantic interpretations if the semantic clarity level is below the predetermined threshold.

4 . The method according to claim 1 , wherein saving the first semantic interpretation is performed independent of a conversation between the user and the AI model comprising data beyond the first and second textual statements.

5 . The method according to claim 1 , further comprising displaying the second plurality of semantic interpretations to the user in a manner wherein the first semantic interpretation is displayed as a first option of the second plurality of semantic interpretations.

6 . The method according to claim 1 , wherein the first plurality of semantic interpretations comprises the first semantic interpretation and a second semantic interpretation, wherein the clarified human intent comprises a first clarified human intent, and wherein the second semantic interpretation comprises a second clarified human intent different than the first clarified human intent.

7 . The method according to claim 1 , wherein the interpretation generation engine comprises at least one of a number of rules, a number of templates, a number of probabilistic models, and at least one machine-learning model.

8 . The method according to claim 1 , wherein generating the AI prompt is further based on the first textual statement.

9 . The method according to claim 1 , wherein the first textual statement is selected from the group consisting of a typed textual statement and a spoken textual statement.

10 . The method according to claim 1 , wherein each of the first plurality of semantic interpretations are framed as a structured query style directive.

11 . A system comprising:

a user input device configured to receive a first textual statement from a user, the first textual statement comprising a first human intent;

an interpretation generation engine configured to receive the first textual statement and employ at least one of a number of optimized AI instructions of the first human intent and a number of trigger structures of the first human intent with the first textual statement in order to generate a first plurality of semantic interpretations of the first textual statement and a first corresponding number of machine-optimized prompt representations for each of the first plurality of semantic interpretations, the first plurality of semantic interpretations and the first corresponding number of machine-optimized prompt representations being generated without exposing the first corresponding number of machine-optimized prompt representations to the user;

a user selection module configured to receive a selection of a first semantic interpretation of the first plurality of semantic interpretations from the user, the first semantic interpretation comprising a clarified human intent and being configured to be saved in a storage after selection from the user;

a prompt formulation engine configured to generate an artificial intelligence (AI) prompt based at least in part on the selection such that the AI prompt is configured to elicit an accurate response from an AI model that is substantially devoid of hallucinations, the response being based on the clarified human intent, the AI prompt being configured to be saved in the storage in order to reduce repeated clarification cost associated with the interpretation generation engine and let the AI prompt compound over time;

an AI engine interface configured to send the AI prompt to the AI model; and

a response handling module configured to generate the response,

wherein the user input device is further configured to receive a second textual statement from the user, the second textual statement comprising a second human intent, and wherein the interpretation generation engine is configured to remember and utilize the selection of the first semantic interpretation to generate a second plurality of semantic interpretations of the second textual statement, the second plurality of semantic interpretations comprising the first semantic interpretation because the second textual statement has a request signature matching a request signature of the first textual statement, thereby allowing the first semantic interpretation to compound over time and wherein the system further comprises a feedback and logging module configured to log data corresponding to the selection, whether clarification on the first plurality of semantic interpretations was requested by the user, and whether the user rated the response as helpful, and refine the interpretation generation engine overtime based on the data.

12 . The system according to claim 11 , wherein the user input device is further configured to allow for selective activation of the interpretation generation engine from the user before receiving the first textual statement from the user.

13 . The system according to claim 11 , wherein the interpretation generation engine is configured to determine whether the first textual statement has a semantic clarity level above a predetermined threshold, and either send the first textual statement directly to the AI model if the semantic clarity level is above the predetermined threshold, or generate the first plurality of semantic interpretations if the semantic clarity level is below the predetermined threshold.

14 . The system according to claim 11 , wherein the first plurality of semantic interpretations comprises the first semantic interpretation and a second semantic interpretation, wherein the clarified human intent comprises a first clarified human intent, and wherein the second semantic interpretation comprises a second clarified human intent different than the first clarified human intent.

Assignments (1)

Continuity (2)

References Cited (45)