Systems and methods for measuring, scoring, and authenticating content generated by artificial intelligence

Abstract

A system and method for generating content from an artificial intelligence (AI) system are disclosed. In one embodiment, the system is trained by one or more training instructors using training content having one or more training content segments, and the method includes accepting first information associating training content with training content attributes, accepting second information associating the AI system with a training instructor profile, generating AI-generated content in response to an AI content generation request from a user, and providing the AI-generated content, the first information, and the second information to the user.

Description

[0001] Cross-references to related applications

[0002] This application claims the benefit of U.S. Provisional Patent Application No. 63 / 546,387, filed November 7, 2023, entitled “System and Method for Measuring, Scoring and Authenticating Artificial Intelligence Produced Contents,” which is incorporated herein by reference. Technical Field

[0003] This disclosure relates to systems and methods for authenticating data and data sources. Background Technology

[0004] Artificial intelligence (AI) is a technology that uses computer-implemented algorithms and databases to perform a variety of tasks. One such task is the generation of content (including media content, software, or other projects). The content generated by AI will soon surpass the content generated by humans.

[0005] AI acquires intelligence through intelligent acquisition. It can acquire intelligence through a simple rule-based paradigm, where a set of rules is "taught" to the AI ​​system. AI can also acquire intelligence through a machine learning-based paradigm. In this paradigm, information is provided to the AI ​​system to find and match patterns, and these patterns are used when generating content and producing information. This learning typically requires a large amount of "training" data for verification. Machine learning is combined with search and optimization techniques, constraint satisfaction, logical and probabilistic reasoning, and control theory. Ultimately, the quality of AI products depends on the quality and selection of the training data, and how well the training data is properly processed to ensure it is accurate, relevant, and of high quality.

[0006] AI systems can be trained using human-generated content, AI-generated content, or a mixture of human and AI-generated content. Furthermore, the instructions used in such training can be benevolent instructions (e.g., reliable instructions intended to produce good results), general instructions (e.g., instructions from general sources), or malicious instructions (e.g., unreliable instructions or instructions intended to produce false results).

[0007] Therefore, AI-generated content may have varying degrees of credibility. Consequently, there is a need to label, measure, and authenticate the nature of such content. For example, recipients of such AI-generated content may want to know the extent to which the AI ​​content was generated purely by humans, by AI operations or content approved by humans, or by AI operations unbeknownst to humans. Recipients of such AI-generated content may also want to know whether the AI ​​was generated from purely human content, a mixture of human and AI content, or purely AI content (e.g., trained using purely human content, a mixture of human and AI content, or purely AI content). Users may also want to know whether all content used to train the AI ​​has been source-verified, approved, and measured. Therefore, there is a need for a system and method to provide measurable, traceable, and verifiable information about the quality and origin of AI-generated content, and to provide authentication mechanisms to ensure that such information is not tampered with. Summary of the Invention

[0008] To address the aforementioned requirements, this document discloses a system and method for generating AI-generated content from an AI system trained by one or more training instructors using training content having one or more training content fragments. In one embodiment, the method includes: receiving first information associating training content with training content attributes; receiving second information associating the AI ​​system with a training instructor profile; generating AI-generated content in response to an AI content generation request from a user; and providing the user with the AI-generated content, the first information, and the second information.

[0009] Another embodiment is a device having a processor and a communication-coupled memory storing processor instructions for performing the aforementioned operations. Yet another embodiment is a device for generating content from an AI system trained by one or more training instructors using training content having one or more training content segments. The device includes: a training input classification module for receiving first information that associates a segment of training content with training content attributes according to a content classification profile; a training instructor profile having second information that associates the AI ​​system with training instructor attributes; an AI system core for generating the content in response to an AI content generation request from a user; and an AI content evaluation module for generating an evaluation of the content based on a content evaluation profile having the first and second information, and for providing the generated content and the evaluation to the user.

[0010] The features, functions, and advantages discussed may be implemented independently in various embodiments of the invention, or may be combined in other embodiments, further details of which can be found in the following description and drawings. Attached Figure Description

[0011] Referring now to the accompanying drawings, where similar reference numerals denote corresponding parts throughout the drawings:

[0012] Figure 1 It is a diagram depicting the development and use of typical AI systems;

[0013] Figure 2 It is a diagram illustrating the process of AI machine learning;

[0014] Figure 3A and 3B This is a diagram illustrating the improved AI system;

[0015] Figure 4 It is a diagram depicting exemplary operations performed by an improved AI system; and

[0016] Figure 5 This is a diagram depicting an exemplary computer system that can be used to implement AI systems. Detailed Implementation

[0017] In the following description, reference is made to the accompanying drawings, which form part of and are illustrated by way of example, several embodiments. It should be understood that other embodiments may be utilized and structural changes may be made without departing from the scope of this disclosure.

[0018] Overview

[0019] Figure 1 This diagram depicts the training and use of a typical AI system 102. Training inputs 104 (e.g., training inputs 104A-104N) and training instructions 112 are selected and modified as needed by a training mentor 106 and provided to the AI ​​system 102. The AI ​​system 102, in training phase 102-T, uses this data to train the algorithms and databases used by the AI ​​system 102. User 110 provides a content generation request 114, and the AI ​​system 102, operating in operation phase 102-O, generates AI-generated content 108. The AI-generated content 108 can be reviewed (by mentor 106 or other entities) to determine whether the AI-generated output 108 meets the expected or desired results. If the AI-generated output 108 does not meet the expected or desired results, the training mentor generates more or different training instructions 112 and / or training inputs 104 to further or retrain the AI ​​system 102. This training process implements AI machine learning and is repeated until the AI ​​system 102 generates AI-generated content 108 acceptable to the training mentor 106.

[0020] Figure 2 This is a diagram that further illustrates an exemplary flow of the AI ​​training process.

[0021] Data Collection and Preprocessing 216: In this stage, unstructured raw data 204 is collected, as shown in box 202. This type of data may include image, text, and audio data, such as voice input. Raw data 204 includes noise and may be inconsistent or repetitive, making it unsuitable for use in AI algorithms.

[0022] To ensure the high quality, reliability, accuracy, and performance of the AI ​​system 102, data preprocessing 216 is performed by the training instructor 106 to improve data quality. During this stage, data analysis and organization (typically performed by the human training instructor 106) are conducted with expertise and insight, including correcting data errors, eliminating overlapping data, removing inconsistent data, and resolving data conflicts. This may include filling in or removing missing values, selecting or removing data attributes, merging existing data attributes, and transforming the raw data 204 into a specified type as needed. The result is preprocessed training data 208. Importantly, the AI ​​system 102 is trained using appropriate training data 104 and training instructions 112, and therefore, the individuals (e.g., the training instructor 106) involved in this process should be trusted. The processed training data 208 is then applied to the AI ​​system in training mode 102-T.

[0023] Test data 210 can also be derived from raw data 204 by training instructor 106 and used to test the AI ​​system in operating mode 102-0. Similar to the training data, test data 210 is processed 208 and analyzed before being applied to AI system 102, including exploring standardized data patterns, data mapping, and extracting data based on exploration and inference. Test data 210 is then applied to the AI ​​system in operating mode 102-0, and the resulting AI-generated content 108 is analyzed by training instructor 106 to ensure accuracy and responsiveness (e.g., the AI-generated content 108 responds to parameters defining content generation request 114), as shown in box 214. This analysis serves as feedback to modify AI model 212, collect additional data, or perform additional preprocessing on the collected data 204.

[0024] Then, the AI ​​system 102 can be used in operation mode 102-O to generate AI-generated content 108 in response to content generation request 114. Training instructor 106 can monitor such AI-generated content 108 and content generation request 114 to ensure accuracy and responsiveness, and further train or retrain the AI ​​system 102 as needed.

[0025] Therefore, the accuracy and usefulness of the AI ​​system 102 largely depend on the quality of the processed training data 208 used to train it, and the quality of the processed training data depends on the quality of selecting the original data 204, preprocessing the original data 204, and managing the human training instructor 106 involved in the training process. Since the original data 204 requires human processing and inspection to ensure data quality for complex AI learning, human error is unavoidable. Damaged and incomplete data, as well as differences from the original data 204, can lead to unintended outputs from the AI ​​system 102. Therefore, during the process of building the model used by the AI ​​system 102, the processed training data 206 must be monitored to verify that the AI-generated content 108 responds to the content generation request 114.

[0026] Although services for collecting and processing data to be used to train AI system 102 are available, the reliability of such data remains a concern. The source of the raw data 204 may be unreliable, or such raw data 204 may be provided for malicious purposes. Therefore, the raw data 204 itself must be analyzed, and the source of such data must be identified, so that a traceable learning data collection environment can be established. Furthermore, the training process is opaque to the user, making AI system 100 largely a black box.

[0027] Figure 3A and 3B The diagram illustrates an improved AI system 300 that enables ratings or other quantitative values ​​to be correlated with AI-generated data. These quantitative values ​​provide information about the AI-generated data, allowing data consumers to determine the quality and credibility of the AI-generated data.

[0028] Figure 4 A diagram illustrating exemplary operations that can be used to generate AI-generated data and certified information about the generation of that AI data is presented. This will be combined with... Figure 3A and 3B Discussion Figure 4 A and 4B.

[0029] Turning Figure 4 As shown in box 402, the improved AI system 300 receives first information that associates one or more pieces of processed data 208 (hereinafter referred to as training content 208) with training content attributes.

[0030] Information that associates training content with training content attributes

[0031] In one embodiment, this first information, which associates training content 208 with training content attributes, is used by the Training Input Classification Module (TICM) 302 as described by... Figure 3A and3B The content classification profile 304 shown is generated from the original AI training input 104 determined by the data.

[0032] TICM 302 accepts raw AI training data input 204 and classifies the raw AI training data input 204 according to the content classification profile 304. The content classification profile 304 defines the training content attributes that will be used to classify and score the raw AI training input data 204, as well as the rules for classifying each piece of training content 208 based on them.

[0033] In one embodiment, the first information definition includes one or more of the following training content attributes:

[0034] Author Approval Score provided by the authors of Training Content 208: The approval score is a measure of the degree to which the authors of Training Content 208 approve the use of the data as Training Content 208.

[0035] Contributor Allocation Score: The contributor allocation score indicates the proportion of contributor content used in generating training content 208. For example, a 70% contributor allocation score could indicate that approximately 30% of training content 208 originates from the author of training content 210, with the remaining 70% from other contributors. In one embodiment, the contributor allocation score includes at least one of the following: a human-created contributor content score (e.g., representing the percentage of human-created content in training content 208), an approved AI-created contributor content score (representing the percentage of training content 208 created by approved AI sources), and an unapproved AI-created contributor training content score (representing the percentage of training content 208 created by unapproved AI sources). For example, all of the aforementioned contributor allocation scores could be expressed as a percentage of the total training content 208. For example, the contributor allocation score for a specific original AI training input might be: 50% human-created contributor content, 40% approved (approved by mentor 106) AI-created contributor content, 5% unapproved AI-created contributor content, and 5% a mixture of approved AI-created content, unapproved AI-created content, and human-created contributor content.

[0036] AI System Signature: This is a digital signature that represents a specific AI system 300 with one or more certified mentor profiles.

[0037] An indication of whether training content 104 has been digitally signed by its author: In one embodiment, the indication is a binary value, where one state indicates that the training content has been digitally signed, and another state indicates that the author of the training content has not digitally signed it. Digitally signed training content (especially by authors known for their data accuracy and reliability) is generally considered more reliable.

[0038] Signature of the author of training content 104: Any element of the first information can be signed by the author of training content 104. For example, training content 104 itself can be hashed and signed by the author of the training content and included in the first information. Additionally, author-approved scores and contributor-allocated scores can also be signed by the author.

[0039] like Figure 3A As indicated in the document, the original AI training content 104 of the first information can also be categorized into the following groups based on credibility, whether the AI ​​training content is signed, and the author (AI or human):

[0040] Training content 208 for AI-generated and signed data with a credibility of A%;

[0041] 208 AI-generated but unsigned training contents with a credibility of B;

[0042] 208 training contents signed and created by AI users;

[0043] 208 unsigned AI user-generated training contents

[0044] AI-generated training content with user signatures and a credibility rating of C% (208);

[0045] AI-generated training content generated by AI users without signatures and with a credibility of D% (208);

[0046] 208 training contents generated by unknown humans with a credibility of E%;

[0047] 208 training contents generated by an unknown AI with a confidence level of F;

[0048] Or any similar classification. This information can be obtained from the metadata provided with the training content 208, or from the human tutor 106 who examines the training content 208 and classifies and scores it.

[0049] return Figure 4 The improved AI system 300 also receives second information that associates the AI ​​system 102 with the training mentor profile 306 and the personal training mentor profile, as shown in boxes 404 and 406.

[0050] Information linking AI systems with aggregated and personal training mentor profiles

[0051] Training mentor 106 possesses a personal training mentor profile, which indicates their experience, qualification level, training quality, feedback from AI users 110, and other factors. This information may include one or more of the following: training mentor certification information (e.g., professional certification, academic qualifications), mentor experience and / or skill scores, mentor professional scores in specific subject areas of the generated content (e.g., mentors may have been given standardized tests in subject areas of interest), and mentor evaluation scores based on reviews from 300 users of one or more other AI systems.

[0052] In addition to or in place of individual tutor profiles, aggregated training tutor profiles can be used. Such aggregated profiles describe the demographics of two or more of the 106 registered training tutors (e.g., the percentage of training tutors 106 who have obtained certification or possess a specific educational or experience qualification). For example, demographics can be represented as a distribution of a specific educational or experience qualification. For instance, individual tutors can be categorized as those who have completed a certification training course or achieved a specific educational status. In this case, the aggregated training tutor profile can present, for example, a graph showing how many tutors in the group of 106 registered training tutors have completed each training course or achieved each educational status. Such aggregated training tutor profiles allow user 100 to obtain AI-generated output 308 generated by AI system 300 to determine whether the group of training tutors 106 as a whole possesses the required experience in relation to the subject matter of AI-generated output 308.

[0053] Similarly, aggregated training instructor profiles may include information describing the experience or skill scores of two or more training instructors on the subject of the training content. This presents, for example, the distribution of skill scores for a group of profiled instructors (e.g., obtained through examination). Aggregated training instructor profiles may also include the distribution of review scores for training instructors, where each review score is based on another person's (other than the reviewed subject's) assessment of the credibility and approval of the generated content.

[0054] return Figure 4 One or more registered mentors 106 train the AI ​​system 300 by generating training instructions, which are used by the AI ​​system core in training phase 102-T to operate on training content 104. This is illustrated in box 408.

[0055] After the AI ​​system 300 is trained, the AI ​​system 300 (running in the operation phase (not the training phase) 102-O) receives a content generation request from the AI ​​user 110. The AI ​​system 300 then generates the requested AI content 318, as shown in box 412.

[0056] In box 414, the generated content, first information (about the training content), and second information (about the tutor profile) are provided to the user 110.

[0057] In box 416, the AI ​​user authentication module 314 uses first information (regarding training content), second information (regarding tutor demographics), and optional AI user 110 preferences to generate third information (evaluated AI output 312) for the AI-generated content 318. The third information includes an evaluation and authentication of the AI-generated content 318. The evaluation may include one or more scores, which may include, for example, an evaluation score (evaluating the usefulness of the AI-generated content) and a credibility score (evaluating the credibility of the generated AI content). In one embodiment, this is... Figure 3B The AI ​​content evaluation module 308 shown is executed using the AI ​​content evaluation profile 310. When evaluating AI-generated content, the AI ​​content evaluation module 308 uses information from the content category profile 304 and the AI ​​tutor profile 306. The preferences of the AI ​​user 110 (e.g., which factors the AI ​​user 110 considers most important when evaluating content, or how these factors should be weighted) can also be input into the AI ​​content evaluation module. In another embodiment, the score is generated by the AI ​​user 110 using the first and second information. In yet another embodiment, the score is generated by the AI ​​content evaluation module 308, but can be modified by the AI ​​user 110. The evaluation score, credibility score, and AI-generated content 318 can be signed by the AI ​​system 300 (specifically, the AI ​​content evaluation module 308).

[0058] Now for reference Figure 4 Box 418 generates AI user authentication output 316. In one embodiment, this includes a user score, a user signature, optional comments, an evaluation score from the evaluated AI output 312, and an AI system signature from the evaluated AI output 312.

[0059] This can be achieved by AI user 110 using AI user authentication module 308 to authenticate and sign the AI-generated output and provide information describing the AI ​​user's assessment of the credibility of the generated content. This can be represented by an evaluation score and a credibility score generated by AI content evaluation module 308, which can be modified by AI user 110 as appropriate. AI user 110 can also use first information and second information in this assessment to evaluate AI-generated content 318 and can sign AI-generated content 318 and / or (multiple) scores. This information is included in AI user authentication output 318, which includes AI-generated content, an evaluation score contributed by AI content evaluation module 308 and optionally modified by AI user 110, and a credibility score. AI user authentication output 316 can also include a score from AI user 110 regarding the AI ​​user's personal evaluation of the AI-generated content, the AI ​​user's signature, and the AI ​​user's comment.

[0060] Finally, refer to Figure 4 Box 420, AI user authentication output 316 is provided as training input for generating second generated AI content.

[0061] Hardware environment

[0062] Figure 5 An exemplary computer system 800 is shown that can be used to implement the processing elements disclosed above, including any processor with computational processing threads. Computer 502 includes one or more processors (CPU, GPU, etc.) 504 and memory (e.g., random access memory (RAM)) 506. Computer 502 is operatively coupled to a display 922, which presents images (such as windows) to a user on a graphical user interface 918B. Computer 902 can be coupled to other devices, such as a keyboard 914, a mouse device 916, a printer 928, etc. Of course, those skilled in the art will recognize that any combination of the above-described components or any number of different components, peripherals, and other devices can be used with computer 902.

[0063] Typically, computer 902 operates under the control of operating system 908 stored in memory 906 and interacts with the user to accept input and commands, presenting results through graphical user interface (GUI) module 518A. Although GUI module 518B is depicted as a separate module, the instructions for performing GUI functions may be retained or distributed within operating system 508, computer program 510, or implemented using dedicated memory and processor. Computer 502 also implements compiler 512, which allows applications 510 written in programming languages ​​such as Python, C / C++, Java, or other languages ​​to be translated into code readable by processor 504. Once completed, application 510 uses the relationships and logic generated using compiler 512 to access and manipulate data stored in memory 506 of computer 502. Computer 502 may also optionally include external communication devices, such as modems, satellite links, Ethernet cards, or other devices for communicating with other computers.

[0064] In one embodiment, the instructions implementing the operating system 508, computer program 510, and compiler 512 are tangibly embodied in a computer-readable medium (e.g., data storage device 520), which may include one or more fixed or removable data storage devices, such as a zip drive, floppy disk drive 524, hard disk drive, CD-ROM drive, magnetic tape drive, etc. Furthermore, the operating system 508 and computer program 510 include instructions that, when read and executed by computer 502, cause computer 502 to perform the operations described herein. Computer program 510 and / or operating instructions may also be tangibly embodied in memory 506 and / or data communication device 530, thereby forming a computer program product or article of manufacture. Therefore, the terms “article of manufacture,” “program storage device,” and “computer program product” as used herein are intended to cover a computer program accessible from any computer-readable device or medium.

[0065] Without departing from the scope of this disclosure, those skilled in the art will recognize that many modifications can be made to this configuration. For example, those skilled in the art will recognize that any combination of the above-described components or any number of different components, peripherals, and other devices can be used.

[0066] in conclusion

[0067] The conclusion includes a description of preferred embodiments of the present disclosure.

[0068] In summary, this document discloses a system and method for generating artificial intelligence (AI) generated content from an AI system trained by one or more training instructors using training content having one or more training content fragments. In one embodiment, the method includes: receiving first information associating training content with training content attributes; receiving second information associating the AI ​​system with a training instructor profile; generating AI-generated content in response to an AI content generation request from a user; and providing the user with the AI-generated content, the first information, and the second information.

[0069] The implementation may include one or more of the following features.

[0070] In any of the above methods, the first information and the second information are signed by the AI ​​system.

[0071] Any of the above methods, wherein the method further comprises: generating third information, the third information including evaluated AI-generated content, the evaluated AI-generated content including AI-generated content evaluation and AI system certification of AI-generated content; and generating user-certified AI-generated content, the user-certified AI-generated content including the AI-generated content, the third information and user score.

[0072] Any of the above methods, wherein: the evaluation includes an evaluation score and a credibility score signed by the AI ​​system.

[0073] In any of the above methods, the user-authenticated AI-generated content is signed by the user.

[0074] In any of the above methods, the user-authenticated AI-generated content is provided as training input for generating the second generated AI content.

[0075] In any of the above methods, wherein: the first information includes one or more of the following: a binary indication of whether the training content is digitally signed by the author of the training content; an author approval score provided by the author of the training content, the author approval score indicating a measure of the author's approval of the training content; a contributor allocation score, the contributor allocation score indicating the proportion of contributor content used in generating the training content; and an AI signature with a certified mentor profile.

[0076] Any of the above methods, wherein: the first information includes the signature of the author of the training content.

[0077] Any of the above methods, wherein the contributor allocation score includes at least one of the following: human-created contributor content; approved AI-created contributor content; unapproved AI-created contributor content; and a mixture of approved AI, unapproved AI, and human-created contributor content.

[0078] In any of the above methods, wherein: the training tutor profile includes an aggregated training tutor profile, the aggregated training tutor profile including at least one of the following: the distribution of training for two or more of the training tutors; the distribution of experience or skill scores for two or more of the training tutors in relation to the topic of the training content; the distribution of review scores for the training tutors, each review score being based on an assessment by another of the credibility of the generated content and the approval of the generated content.

[0079] In any of the above methods, the training mentor profile includes a personal profile of the one or more mentors, the personal profile including one or more of the following: mentor certification information; mentor experience score; mentor skill score; mentor professional score in relation to the topic of the generated content; and mentor review score from another AI system user.

[0080] Another embodiment is a device for generating content from an artificial intelligence (AI) system trained by one or more training instructors using training content having one or more training content segments. In one embodiment, the device includes: a processor; and a memory communicatively coupled to the processor, the memory storing processor instructions, the processor instructions including processor instructions for performing the following operations: receiving first information associating a segment of training content with training content attributes; receiving second information associating the AI ​​system with a training instructor profile; generating the content in response to an AI content generation request from a user; and providing the generated content, the first information, and the second information to the user.

[0081] The implementation may include one or more of the following features.

[0082] Any of the above-mentioned devices, wherein: the first information and the second information are signed by the AI ​​system.

[0083] Any of the aforementioned devices, wherein: the processor instructions further include processor instructions for performing the following operations: generating third information, the third information including evaluated AI-generated content, the evaluated AI-generated content including AI-generated content evaluation and AI system certification of the AI-generated content; and generating user-certified AI-generated content, the user-certified AI-generated content including the AI-generated content, the third information, and a user score.

[0084] Any of the above-mentioned devices, wherein: the evaluation includes a user rating score and a credibility score; and the user-authenticated AI-generated content is signed by the user.

[0085] Any of the above-mentioned devices, wherein: the user-authenticated AI-generated content is provided as training input for generating the second generated AI content.

[0086] Any of the above-mentioned devices, wherein: the first information includes one or more of the following: a binary indication of whether the training content is digitally signed by the author of the training content; an author approval score, provided by the author of the training content, the author approval score indicating a measure of the author's approval of the training content; a contributor allocation score, the contributor allocation score indicating the proportion of contributor content used in generating the training content; and an AI signature with a certified mentor profile.

[0087] Any of the aforementioned devices, wherein: the contributor allocation score includes at least one of the following: human-created contributor content; approved AI-created contributor content; unapproved AI-created contributor content; and a mixture of approved AI, unapproved AI, and human-created contributor content.

[0088] Any of the above-mentioned devices, wherein: the training instructor profile includes an aggregated training instructor profile, the aggregated training instructor profile including at least one of the following: the distribution of training for two or more of the training instructors; the distribution of experience or skill scores of two or more of the training instructors in relation to the topic of the training content; the distribution of review scores of the training instructors, each review score being based on an assessment by another of the credibility of the generated content and the approval of the generated content.

[0089] Another embodiment is a device for generating content from an artificial intelligence (AI) system trained by one or more training instructors using training content having one or more training content segments. In one embodiment, the device includes: a training input classification module for receiving first information that associates a segment of training content with training content attributes according to a content classification profile; a training instructor profile having second information that associates the AI ​​system with training instructor attributes; an AI system core for generating the content in response to an AI content generation request from a user; and an AI content evaluation module for generating an evaluation of the content based on a content evaluation profile having the first and second information, and for providing the generated content and the evaluation to the user.

[0090] For purposes of illustration and description, the foregoing description of preferred embodiments has been presented. This is not intended to be exhaustive or to limit the present disclosure to the precise forms disclosed. In view of the foregoing teachings, many modifications and variations are possible. The scope of the claims is not limited by this detailed description but by the claims appended herein.

Claims

1. A method for generating AI-generated content from an artificial intelligence (AI) system, said AI system being trained by one or more training instructors using training content having one or more training content fragments, said method comprising: Receive the first information that associates the training content with the attributes of the training content; Accept second information that associates the AI ​​system with a training instructor profile; The AI-generated content is generated in response to an AI content generation request from a user; The user is provided with the AI-generated content, the first information, and the second information.

2. The method according to claim 1, wherein the first information and the second information are signed by the AI ​​system.

3. The method according to claim 2, further comprising: Generate third information, which includes evaluated AI-generated content, comprising an AI-generated content assessment and an AI system certification of the AI-generated content; and Generate AI-generated content that has been authenticated by the user. The AI-generated content that has been authenticated by the user includes the AI-generated content, the third-party information, and the user score.

4. The method of claim 3, wherein the evaluation includes an evaluation score and a credibility score signed by the AI ​​system.

5. The method of claim 4, wherein the user-authenticated AI-generated content is signed by the user.

6. The method of claim 5, wherein the user-authenticated AI-generated content is provided as training input for generating the second generated AI content.

7. The method of claim 1, wherein the first information comprises one or more of the following: A binary indication of whether the training content is digitally signed by the author of the training content; Author Approval Score, provided by the authors of the training content, indicates a measure of the authors' approval of the training content. Contributor allocation score, which indicates the proportion of contributor content used when generating the training content; as well as AI signature with certified mentor profile.

8. The method of claim 7, wherein the first information includes the signature of the author of the training content.

9. The method of claim 7, wherein the contributor score allocation includes at least one of the following: Content contributed by human creators; Approved AI-generated contributor content; Unauthorized AI-generated contributor content; and A mix of approved AI, unapproved AI, and human-created contributor content.

10. The method of claim 7, wherein the training tutor profile includes an aggregated training tutor profile, the aggregated training tutor profile including at least one of the following: The distribution of training among two or more of the training instructors; The distribution of experience or skill scores of two or more of the training instructors in relation to the subject of the training content; The distribution of the training instructors' review scores, with each review score based on an assessment by another person of the credibility and approval of the generated content.

11. The method of claim 10, wherein the training instructor profile includes personal profiles of the one or more instructors, the personal profiles including one or more of the following: Instructor certification information; Mentor experience score; Mentor skills score; The tutor's professional score in relation to the topic of the generated content; and The mentor review score from another AI system user.

12. An apparatus for generating content from an artificial intelligence (AI) system, said AI system being trained by one or more training instructors using training content having one or more training content segments, said apparatus comprising: processor; A memory communicatively coupled to the processor, the memory storing processor instructions, the processor instructions including processor instructions for performing the following operations: Receive the first information that associates a piece of training content with the attributes of the training content; Accept second information that associates the AI ​​system with a training instructor profile; The content is generated in response to an AI content generation request from a user; and The generated content, the first information, and the second information are provided to the user.

13. The device of claim 12, wherein the first information and the second information are signed by the AI ​​system.

14. The apparatus of claim 13, wherein the processor instructions further include processor instructions for performing the following operations: Generate third information, which includes evaluated AI-generated content, comprising an AI-generated content assessment and an AI system certification of the AI-generated content; and Generate AI-generated content that has been authenticated by the user. The AI-generated content that has been authenticated by the user includes the AI-generated content, the third-party information, and the user score.

15. The device according to claim 14, wherein: The evaluation includes user rating scores and credibility scores; and AI-generated content that has been authenticated by the user is signed by the user.

16. The device of claim 15, wherein the user-authenticated AI-generated content is provided as training input for generating the second generated AI content.

17. The device of claim 12, wherein the first information includes one or more of the following: A binary indication of whether the training content is digitally signed by the author of the training content; Author Approval Score, provided by the authors of the training content, indicates a measure of the authors' approval of the training content. Contributor allocation score, which indicates the proportion of contributor content used when generating the training content; as well as AI signature with certified mentor profile.

18. The device of claim 17, wherein the contributor allocation score includes at least one of the following: Content contributed by human creators; Approved AI-generated contributor content; Unauthorized AI-generated contributor content; and A mix of approved AI, unapproved AI, and human-created contributor content.

19. The device of claim 12, wherein the training instructor profile includes an aggregated training instructor profile, the aggregated training instructor profile including at least one of the following: The distribution of training among two or more of the training instructors; The distribution of experience or skill scores of two or more of the training instructors in relation to the subject of the training content; The distribution of the training instructors' review scores, with each review score based on an assessment by another person of the credibility and approval of the generated content.

20. An apparatus for generating content from an artificial intelligence (AI) system, said AI system being trained by one or more training instructors using training content having one or more training content segments, said apparatus comprising: A training input classification module is used to receive first information that associates a piece of training content with training content attributes according to a content classification profile. A training mentor profile, which includes second information that associates the AI ​​system with the attributes of the training mentor. The core of the AI ​​system is used to generate the content in response to an AI content generation request from a user; and The AI ​​content evaluation module is used to generate an evaluation of the content based on a content evaluation profile containing the first information and the second information, and to provide the generated content and the evaluation to the user.

Images