Server system and program

Abstract

To provide a new technology that can reduce the processing load on a server system that provides users with content including content elements generated by a generation unit. [Solution] A server system that controls the provision of content to a user terminal distributes the content generation units responsible for generating content elements that constitute a given target content between a server-side generation unit and a terminal-side generation unit on the user terminal. The server system causes each responsible generation unit to generate the content elements it is responsible for, controls the progress of the target content including the generated content elements, and provides the target content to the user terminal.

Description

Technical Field

[0001] The present invention relates to a server system and the like.

Background Art

[0002] In recent years, AI (Artificial Intelligence) has been increasingly used in games. Patent Document 1 describes a technique for providing recommendations to players using an AI model.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] As one form of using AI in video games, it is conceivable to have a generation unit represented by generative AI generate content elements (for example, scenarios, game stages, stage maps, characters that appear, items that appear, backgrounds, etc.) that make up the game content.

[0005] Generative AI can be roughly divided into generative AI installed and implemented in a server system and generative AI installed and implemented in a user terminal such as a smartphone. The generative AI installed and implemented in a user terminal is called an "AI app" or the like.

[0006] [[ID=4)]]

[0007] These problems are not limited to video games; they also apply to content that provides players with virtual experiences in virtual spaces, when using generative AI (generative components).

[0008] The problem that this invention aims to solve is to provide a new technology that can reduce the processing load on a server system that provides users with content including content elements generated by a generation unit. [Means for solving the problem]

[0009] The first invention for solving the above problems is a server system that controls the progress of content and provides said content to a user terminal, The user terminal has the functions of a terminal-side generation unit (for example, terminal-side generation AI5 in Figure 1, terminal-side generation unit 280 in Figure 14), A generation planning means (for example, the generation planning unit 222 in Figure 9, step S22 in Figure 15) selects whether the generation unit responsible for generating content elements constituting a given target content is a server-side generation unit or a terminal-side generation unit, A generation control means (for example, the generation control unit 224 in Figure 9, steps S148 and S152 in Figure 16) that controls the assigned generation unit to generate the content elements it is responsible for, The server system includes a content provision control means (for example, the content provision control unit 226 in Figure 9, step S172 in Figure 16) that controls the progress of the target content, which includes the content elements generated by the responsible generation unit, and controls the provision of the target content to the user terminal.

[0010] The second invention is a server system in which, in the above-mentioned server system, there are multiple content elements, and the generation planning means distributes the content elements to a server-side generation unit and a terminal-side generation unit, assigning each content element to a corresponding generation unit.

[0011] The third invention is a server system in which, in the above-described server system, the content provision control means performs control to aggregate the generated plurality of content elements and provide the target content (for example, steps S148 and S152 in Figure 16).

[0012] According to the first to third inventions, the server system can distribute the generation of content elements constituting the target content between a server-side generation unit, which is part of the server system itself, and a terminal-side generation unit, which is part of the user terminal. Therefore, it becomes possible to reduce the processing load on the server system that provides the user with content including the content elements generated by the generation units.

[0013] The fourth invention is a server system in which, in the above-described server system, the generation planning means selects the responsible generation unit based on the content elements (for example, step S100 in Figure 19).

[0014] According to the fourth invention, the server system can allocate the appropriate generation unit to the appropriate location based on the content elements, thereby more effectively reducing the processing load on the server system.

[0015] The fifth invention is a server system in which, in the above-described server system, the generation planning means selects the responsible generation unit based on the degree to which the content element influences the progress of the target content (for example, event definition data 514 in Figure 8, step S70 in Figure 18).

[0016] According to the fifth invention, the server system can select a responsible generation unit based on the degree to which a content element influences the progress of the target content.

[0017] The sixth invention is a server system further comprising a prediction means (for example, the prediction unit 220 in Figure 9, step S44 in Figure 17) for predicting the generation cost related to the generation of the content elements, wherein the generation planning means selects the responsible generation unit based on the predicted generation cost (for example, step S74 in Figure 18, the second predicted generation cost sequential score in step S102 in Figure 19).

[0018] The seventh invention is a server system in which, in the above-described server system, the predicted generation cost includes predicted load information related to the generation of the content elements (for example, predicted load information 562 in Figure 7), and the generation planning means selects the responsible generation unit based on the predicted load information.

[0019] According to the sixth or seventh invention, the server system becomes capable of selecting a responsible generation unit based on the predicted generation cost.

[0020] The eighth invention is a server system in which, in the above-described server system, the generation planning means selects the responsible generation unit based on user information (for example, step S106 in Figure 19).

[0021] According to the eighth invention, the server system becomes capable of selecting a responsible generation unit based on user information.

[0022] The ninth invention is a server system in which, in the above-described server system, the user information includes privacy information, and the generation planning means determines whether or not the content element is an element that is generated using the privacy information, and if the determination is affirmative, selects the terminal-side generation unit as the responsible generation unit for the content element (for example, step S72 in Figure 18).

[0023] According to the ninth invention, the server system can select a terminal-side generation unit as the responsible generation unit when the content element is an element generated using the aforementioned privacy information.

[0024] The tenth invention is a server system which further includes terminal information acquisition control means (for example, the terminal information acquisition control unit 210 in FIG. 9, step S12 in FIG. 15) for acquiring terminal information from the user terminal, and the generation planning means selects the responsible generation unit based on the terminal information (for example, the processing capacity ranking points in step S102 in FIG. 19).

[0025] According to the tenth invention, the server system can select the responsible generation unit based on the terminal information from the user terminal.

[0026] The eleventh invention is a server system which further includes content play information acquisition control means (for example, the content play information acquisition control unit 214 in FIG. 9, step S104 in FIG. 19) for acquiring content play information related to the user terminal, and the generation planning means selects the responsible generation unit based on the content play information.

[0027] According to the eleventh invention, the server system can select the responsible generation unit based on the content play information

[0028] The twelfth invention is a server system which further includes terminal side generation status information acquisition control means (for example, the terminal side generation status information acquisition control unit 212 in FIG. 9, step S14 in FIG. 15) for controlling the acquisition of terminal side generation status information indicating the generation status of the terminal side generation unit from the user terminal, and the generation planning means selects the responsible generation unit based at least on the terminal side generation status information (for example, the terminal side generation status ranking points in step S102 in FIG. 19).

[0029] According to the twelfth invention, the server system can select the responsible generation unit based on the generation status of the terminal side generation unit.

[0030] The 13th invention is a server system further comprising: generation planning means (for example, generation planning unit 216 in Figure 9, step S18 in Figure 15) that performs control to plan a generation instruction that causes the generation of the content element; and terminal-side planning information acquisition control means (for example, terminal-side planning information acquisition control unit 218 in Figure 9, step S18 in Figure 15) that acquires terminal-side planning information that plans the generation instruction from the user terminal, wherein the generation planning means selects the responsible generation unit (for example, step S76 in Figure 18) based on whether the generation instruction is from the generation planning means or based on the terminal-side planning information.

[0031] According to the 13th invention, the server system can select a responsible generation unit based on whether or not the generation instruction is based on terminal-side intention information.

[0032] The fourteenth invention is a server system in which, in the above-described server system, the content element includes replaceable content elements that can be replaced during the progress control of the content (for example, the replaceable content element list 512 in Figure 8), the generation planning means selects the terminal-side generation unit as the responsible generation unit for the replaceable content elements, and the content provision control means performs control to replace the content element using the replaceable content elements during the progress control of the content (for example, step S172 in Figure 16).

[0033] According to the 14th invention, the server system becomes capable of replacing content elements using replaceable content elements during content progression control.

[0034] The fifteenth invention is a server system in which, in the above-described server system, the generation control means creates generation instruction information for generating the content elements and provides it to the responsible generation unit (for example, step S120 in Figure 18).

[0035] According to the 15th invention, the server system can create generation instruction information for the assigned generation unit.

[0036] The sixteenth invention is a server system further comprising a past generation information storage means (for example, the past generation information storage control unit 228 in Figure 9, the IC memory 1152, and step S160 in Figure 16) that stores the generated content elements as past generation information in association with the information at the time of generation.

[0037] According to the 16th invention, the server system can store the generated content elements as previously generated information, associated with the information at the time of generation.

[0038] The 17th invention is a server system in which, in the above-described server system, the generated information includes at least one of the following: information on the type and / or quality of the content elements and information on the generation cost (for example, past generation information 780 in Figure 8), the generation planning means decides, based on the generated information, to reuse the generated content elements included in the past generation information instead of selecting the responsible generation unit (for example, step S48 in Figure 17, substitute content element ID 765 in Figure 12), and the content provision control means controls the provision of the target content using the generated content elements included in the past generation information when the reuse is decided by the generation planning means (for example, step S144 in Figure 15).

[0039] According to the 17th invention, the server system can reuse previously generated content elements included in past generated information instead of selecting a responsible generation unit.

[0040] The 18th invention is a program for a user terminal to communicate with a server system to control the progress of content, comprising: a generation planning means for selecting whether the responsible generation unit responsible for generating content elements constituting a given target content is a server-side generation unit of the server system or a terminal-side generation unit; a generation control means for controlling the responsible generation unit I to generate the content elements it is responsible for; and a content progress control means for controlling the progress of the target content including the content elements generated by the responsible generation unit.

[0041] According to the 18th invention, a program can be realized that allows a user terminal to perform the same functions as in the first invention. [Brief explanation of the drawing]

[0042] [Figure 1] A system configuration diagram showing an example of a content delivery system. [Figure 2] A diagram to explain the origin (cause) of new content creation. [Figure 3] A diagram illustrating an example of a user interface. [Figure 4] A diagram illustrating an example of a user interface. [Figure 5] A conceptual diagram to explain generative planning. [Figure 6] A diagram illustrating generative planning. [Figure 7] A diagram showing an example of the data structure for prediction pattern data. [Figure 8] A diagram showing examples of programs and data stored by a server system. [Figure 9] A diagram showing an example of the functional configuration of the server processing unit. [Figure 10] A diagram showing an example of the data structure of user registration data. [Figure 11] A diagram showing an example of the data structure of AI-generated registration data. [Figure 12]A diagram showing an example of the data structure of the generation planning data. [Figure 13] A diagram showing examples of programs and data stored on a user terminal. [Figure 14] A diagram showing an example of the functional configuration of the terminal processing unit. [Figure 15] A flowchart illustrating the processing flow related to the generation and provision of target content executed by the server system. [Figure 16] Flowchart continuing from Figure 15. [Figure 17] A flowchart illustrating the flow of the generation planning process. [Figure 18] Flowchart continuing from Figure 17. [Figure 19] A flowchart illustrating the process of setting up the AI ​​generated by the terminal side. [Figure 20] A diagram illustrating a modified example. [Figure 21] A diagram illustrating an example of the programs and data stored in a modified server system. [Figure 22] A diagram illustrating an example of the functional configuration of the server processing unit in a modified example. [Figure 23] A diagram illustrating an example of the programs and data stored in a modified user terminal. [Figure 24] A diagram showing an example of the functional configuration of the terminal processing unit in a modified example. [Modes for carrying out the invention]

[0043] Examples of embodiments of the present invention will be described below, but it goes without saying that the embodiments to which the present invention can be applied are not limited to the following embodiments.

[0044] Figure 1 is a system configuration diagram showing an example of the configuration of a content provision system according to this embodiment. The content provision system 1000 is a computer system that provides content offering virtual experiences in a virtual space, such as video games, virtual activities, and shopping, to multiple registered users in parallel. In other words, the content provision system 1000 is able to constantly provide users with new virtual experiences.

[0045] The content provision system 1000 is a computer system that includes a server system 1100 and user terminals 1500 for each user, all connected via a network 9 for data communication.

[0046] Network 9 refers to a communication path capable of data transmission. In other words, Network 9 includes not only LANs (Local Area Networks) using dedicated lines (dedicated cables) or Ethernet (registered trademark) for direct connections, but also telephone networks, cable networks, and the Internet.

[0047] The server system 1100 is a computer system that performs various processes such as managing and controlling registered user information and various controls related to content provision (for example, controlling the progress of a game).

[0048] The server system 1100 has a control board 1150 mounted on the main unit 1101. The control board 1150 is equipped with various microprocessors such as a CPU (Central Processing Unit) 1151, a GPU (Graphics Processing Unit), and a DSP (Digital Signal Processor), various IC memories 1152 such as VRAM, RAM, and ROM, and a communication device 1153. Some or all of the functions mounted on the control board 1150 may be implemented using an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array), or a SoC (System on a Chip).

[0049] Although the server system 1100 is depicted as a single server device, it may also be configured with multiple devices. For example, the server system 1100 may be configured with multiple servers, each handling a different function, and connected to each other via an internal bus or network 9 for data communication. Furthermore, the server system 1100 may include a database and online storage.

[0050] The server system 1100 is equipped with a server-side generation AI 3 that generates content based on given generation instruction information (e.g., prompts). The server-side generation AI 3 is implemented using a machine learning-based AI model on hardware employing a multi-core architecture (e.g., a group of GPUs and memory, a group of AI chips, etc.). The hardware for the server-side generation AI 3 is not limited to dedicated hardware.

[0051] The server-side generation AI3 may be implemented as a single multimodal generation AI model, or it may be configured as a group of AIs having one or more element-specific server-side generation AIs 4 (4a, 4b, ...) for each type of content element. Hereafter, we will describe the case where there are multiple element-specific server-side generation AIs 4.

[0052] Content elements can include, for example, if the content is a video game, the game's scenario, game stage maps, characters appearing in the game, character dialogue and actions, etc. Other examples include the game's background music, background images, background objects, etc.

[0053] For example, if the content is a virtual underwater dive experience (one of the virtual activities), the types of content elements may include the diving space (stage), the types of creatures that appear in the water, and artificial objects that appear in the water (e.g., ships, submarines, shipwrecks, etc.). The server-side generated AI3 may be configured to have separate server-side generated AI4 for each of these content elements.

[0054] There may be at least one element-specific server-side generation AI4 for each type of content element. The types of content elements may be further subdivided, and a configuration with element-specific server-side generation AI4 may be provided for each subdivision. Specifically, a configuration may include an element-specific server-side generation AI4 that generates monster-like characters, an element-specific server-side generation AI4 that generates cute, pop-style characters, or an element-specific server-side generation AI4 that generates robot-like characters, etc.

[0055] User terminal 1500 serves as a Man-Machine Interface (MMIF) for user 2 to play content as a player. For example, if the content to be played is a game, the user terminal 1500 for user 2, who is the player, functions as a game playing terminal. Although only one user terminal 1500 is depicted in Figure 1, in actual operation, it is common for multiple user terminals 1500 to communicate and connect to the server system 1100 simultaneously.

[0056] The user terminal 1500 is a computer system that can connect to the network 9, such as a personal computer, smartphone, wearable computer, portable game console, home game console, or tablet computer.

[0057] The user terminal 1500 is a computer comprising an operation input device, an image display device, a communication device, and a control board 1550 that performs calculation processing. Examples of the operation input device include a touch panel 1506, a keyboard, a game controller, and a mouse. Examples of the image display device include a touch panel 1506, a head-mounted display, and a glasses-type display.

[0058] The control board 1550 is equipped with a CPU 1551, various microprocessors such as a GPU and DSP, various IC memories 1552 such as VRAM, RAM, and ROM, and a communication module 1553 that connects to the network 9. These elements mounted on the control board 1550 are electrically connected via bus circuits and the like, enabling data reading and writing, and signal transmission and reception. Part or all of the control board 1550 may be an ASIC, FPGA, or SoC.

[0059] The control board 1550 stores programs and various data necessary to realize the functions of the user terminal 1500 in the IC memory 1552. The user terminal 1500 realizes its functions as a play terminal by executing a predetermined application program.

[0060] The user terminal 1500 is equipped with a terminal-side generation AI 5 capable of generating content based on given generation instruction information. The terminal-side generation AI 5 may be implemented as an application or as hardware as an AI chip.

[0061] The terminal-side generation AI 5 of the user terminal 1500 may be implemented as a single multimodal generation AI model, or it may be configured as a group of AIs having one or more element-specific terminal-side generation AIs 6 (6a, 6b, ...) for each type of content element. There may be at least one element-specific terminal-side generation AI 6 for each type of content element. The types of content elements may be further classified, and a configuration with element-specific terminal-side generation AI 6 for each classification may be used. Hereafter, the terminal-side generation AI 5 will be described as having multiple element-specific terminal-side generation AIs 6.

[0062] Figure 2 is a diagram illustrating the origin (cause) of the creation of new content. The creation of new content originates from the intentions of either server system 1100 or user 2, who is a player.

[0063] For example, during game progress control (while providing content), the server system 1100 attempts to generate target content in events such as the appearance of a new game stage or the addition of a new character. For example, during gameplay (while playing content), user 2 attempts to generate target content, such as wanting to play a new game stage or wanting to add a new character, by performing a predetermined target content request operation.

[0064] Figure 3 shows an example of a user interface related to the target content request operation when the target content is intended to be a "new game stage". Figure 4 shows an example of a user interface related to a target content request operation when "a new character" is intended as the target content.

[0065] When a target content request operation is performed, a category setting UI (User Interface) 10 is displayed, which accepts the user 2's selection of the target content category. The number and classification of content categories are not limited to the examples shown in Figures 3 and 4.

[0066] When User 2 selects one of the multiple categories shown in Category Setting UI 10, one or more detailed setting UIs 12 (12a, 12b, ...) are displayed. The detailed settings UI12 is displayed for each detailed setting item related to the content of the selected category, and accepts selection operations from the user. The content and number of items in the detailed settings UI12 are predetermined for each category.

[0067] The detailed settings UI12 includes a privacy information usage settings UI12p (see, for example, Figure 4) which asks User 2 whether it is acceptable to use the privacy information entered by User 2 during user registration to generate the target content.

[0068] Privacy information includes real name, facial image, gender, age, address, user account, etc.

[0069] For example, in the privacy information usage setting UI12p in Figure 4, if the target content is "New Character," it asks whether it is acceptable to set the user's face image or a caricature based on that face image as the character's face. In addition, if the target content is "New Character," a detailed setting UI12 may be provided that asks whether it is acceptable to match the gender of user 2 with the gender of the character to be generated. A detailed setting UI12 may also be provided that asks whether it is acceptable to render the real name or account name as text in the character's design.

[0070] In the display of the privacy information usage settings UI12p, where in the content privacy information is used is pre-configured according to the categories selected in the category settings UI10.

[0071] For example, if the categories are "New Game Stage," "New Item," and "New Technique," the Privacy Information Usage Settings UI12p could ask whether it's okay to include the player's real name in the stage name. For "New Game Stage," it could also ask whether it's okay to include the player's face image in the background.

[0072] Furthermore, the user interface for request operations may not be an icon-based system where users select from pre-defined options, as shown in Figures 3 and 4, but rather a system where user 2 inputs text using natural language. In this case, a separate natural language recognition AI will be provided to recognize from the input text whether the user selected an option from the category setting UI 10 or the detailed setting UI 12.

[0073] Returning to Figure 2, if the entity responsible for the configuration is the server system 1100 (computer), the category setting UI 10 and detailed setting UI 12 are not displayed, and the server system 1100 automatically configures them. For example, the categories may be predetermined according to the events that occur, and the items in the detailed settings may be set randomly.

[0074] If the intention is made by user 2, the user terminal 1500 creates a given terminal-side intention information 832 in response to the input of the target content request operation and sends it to the server system 1100 along with a predetermined request signal. The server system 1100 acquires this information.

[0075] The terminal-side intent information 832 includes various information necessary for creating the generation instruction information. For example, it includes a user account as an identifier of the intent entity, a content category (indicating what kind of information to be generated), generation conditions, required quality, supplementary information, etc. If the intent specifies existing content to be used as a model or source for modification when generating the target content, the data of that existing content and the search conditions for that existing content are considered supplementary information.

[0076] The server system 1100 sets "cause-generating instruction information" based on the intent related to the target content. The cause-generating instruction information 702 includes various information necessary to create the generation instruction information. For example, it includes an automatically assigned intent ID, intent entity identification, content category (indicating what kind of information to generate), generation conditions, required quality, supplementary information, etc. If the intent specifies existing content to be used as a model or source for modification when generating the target content, the data of that existing content and the search conditions for that existing content are used as supplementary information.

[0077] Figure 5 is a conceptual diagram illustrating generative planning. Generation planning is a function designed to reduce the processing load on the generation AI of the server system 1100. Generation planning identifies the types of content elements (the types of elements that make up the target content) that are necessary when generating the target content. Then, for each identified content element, it is decided which generation AI—the element-specific server-side generation AI 4 of the server-side generation AI 3, or the element-specific terminal-side generation AI 6 of the terminal-side generation AI 5—should generate it.

[0078] For example, if the target content is a "game stage," then through generation planning, content elements such as "map," "background," "characters," "visual effects," "sound effects," and "items" are identified. Then, a generation AI is assigned to each content element. The assigned generation AI is called the "assigned generation AI."

[0079] By distributing the responsibility for content elements between the server system 1100 and the user terminal 1500, it becomes possible to reduce the processing load on the server system 1100.

[0080] Figure 6 is a diagram illustrating generative planning. When executing generation planning, the server system 1100 refers to the originating generation instruction information, user information, generation status information of the server system 1100 (server-side generation status information), and generation status information of the user terminal 1500 (terminal-side generation status information).

[0081] User information includes device information, content play information, and privacy information.

[0082] Terminal information refers to various pieces of information indicating the processing capabilities of user terminal 1500 as a computer. For example, this includes CPU model information, memory capacity, communication module specifications, and the types of terminal-generated AI5 and element-specific terminal-generated AI6 (each corresponding to information on the content element; corresponding modality).

[0083] User information includes information on the game progress of the game content being played by the user. From the game progress information, the load on the user's terminal 1500 can be estimated. For example, if the user is playing an action game, a shooting game, or a similar game stage in a game space composed of 3DCG where the player character moves at high speed, it can be estimated that the user's terminal 1500 is under a high load.

[0084] Server-side generation status information includes information indicating the processing load status and request congestion related to server-side generated AI3 and element-specific server-side generated AI4. "Processing load status information" includes, for example, CPU load rate and memory consumption rate (usage rate). "Information indicating congestion status" includes, for example, the number of pending executions.

[0085] Terminal-side generation status information includes information on processing load status related to terminal-side generation AI5 and element-specific terminal-side generation AI6, request congestion status, and factors that affect generation costs. "Processing load status information" includes, for example, CPU load rate and memory consumption rate (usage rate). "Information indicating congestion status" includes, for example, the number of pending executions. "Factors that affect generation costs" include, for example, communication quality (communication speed, PING value, JITTER value, etc.).

[0086] Terminal-side generation status information is constantly monitored for every 1500 user terminals. The server system 1100 can obtain information indicating the processing load status and congestion status managed by the user terminal 1500 from the user terminal 1500 through functions implemented by the client program executed on the user terminal 1500.

[0087] The server system 1100, as part of its generation planning, lists the content elements that constitute the content category indicated by the originating generation instruction information. This is called "content element decomposition." The content elements required for each category of content are defined in advance.

[0088] The server system 1100 predicts the generation cost for each content element as a predicted generation cost. The predicted generation cost is, for example, the time required (predicted generation time) if the content element were to be generated by the element-specific server-side generation AI 4, and the cost required for that. An example of the cost is a usage-based fee for renting the generation AI, which is determined by the number of tokens in the generation instruction information and the size of the data generated.

[0089] Furthermore, the prediction generation cost may not be a single value, but rather a range of values, or it may be categorized as "heavy" or "light."

[0090] Figure 7 shows an example of the data structure for prediction pattern data 540. The prediction pattern data 540 is the data that the server system 1100 uses to predict generation costs, and is prepared for various situations. Each prediction pattern data 540 stores the application situation requirements 542 and the predicted generation cost 560 in association.

[0091] The applicability requirement 542 stores various data that define the conditions that must be met for the situation to be determined to be applicable to the prediction pattern. The applicability requirement 542 is written, for example, by combining multiple subconditions with AND or OR.

[0092] Sub-conditions of the application status requirement 542 include, for example, category conditions 543 regarding the category of the target content, element type conditions 544 regarding the types of content elements, generation condition conditions 545 regarding the generation conditions, required quality conditions 546 regarding the required quality, terminal information conditions 547, communication environment conditions 548, and so on.

[0093] The generation condition 545 is provided for each item of the generation condition and is described as a range of numerical values ​​or thresholds that define the generation condition. The required quality conditions 546 are provided for each required quality item and are described as numerical ranges or thresholds that define the required quality.

[0094] In other words, server-side application requirement 542 specifies "the situation in which content elements of what purpose content are generated, under what generation conditions and required quality."

[0095] Terminal information condition 547 is a condition concerning information that can be used to estimate the level of generation capability. Terminal information condition 547 describes conditions such as CPU model, clock speed, number of cores, amount of installed memory, and whether the generation AI is implemented in software or in hardware such as an AI chip.

[0096] The communication environment conditions 548 are described as the conditions of the communication environment between the user terminal 1500 and the server system 1100. For example, they are described using ranges or thresholds for communication speed, PING value, JITTER value, etc. The communication speed may be obtained by calculating the communication speed from the history of past communications between the server system 1100 and the user terminal 1500.

[0097] Note that terminal information condition 547 and communication environment condition 548 are used as application status requirement 542 for prediction patterns that are assumed to be generated on the user terminal 1500, but are not used for prediction patterns that are assumed to be generated on the server system 1100. The sub-conditions of application status requirement 542 may all be set to effectively "none" or "unlimited".

[0098] The predicted generation cost 560 includes predicted load information 562 and predicted generation fee 564. The predicted load information 562 is information that directly or indirectly indicates the load that generation in the predicted pattern places on the server system 1100 and user terminal 1500. For example, the predicted load information 562 may include the time required for predicted generation and the amount of data to be generated. The predicted generation fee 564 is set when there is a fee for using the generation AI. For example, when using the generation AI in a subscription contract or pay-as-you-go contract where there is an upper limit on the amount of data to be generated by the generation AI, it indicates the fee required to perform generation in the predicted pattern.

[0099] The server system 1100 searches for predictive pattern data 540 that meet the requirements from among the predictive pattern data 540 for application status requirements 542 where terminal information conditions 547 and communication environment conditions 548 are not set. The predictive generation cost 560 of the retrieved predictive pattern data 540 is defined as the "first predictive generation cost" assuming that the generation was performed by the server system 1100.

[0100] Furthermore, the server system 1100 searches for predictive pattern data 540 that meet the requirements from among the predictive pattern data 540 of the application status requirement 542, for which terminal information conditions 547 and communication environment conditions 548 are set. The predictive generation cost 560 of the retrieved predictive pattern data 540 is defined as the "second predictive generation cost" assuming that the generation was performed on the user terminal 1500.

[0101] Returning to Figure 6, the server system 1100, in generation planning, identifies content elements whose first predicted generation cost satisfies a given high-cost condition. Then, among the element-specific server-side generation AIs 4 of the server system 1100, the generation AI capable of generating the content element in question is set as the "assigned generation AI," and element-specific generation instruction information is generated for the assigned generation AI.

[0102] For content elements whose first predicted generation cost does not meet the given high-cost condition, one of the user terminals 1500 whose second predicted generation cost meets the given low-cost condition is selected as the "assigned terminal." Then, from among the element-specific terminal-side generation AIs 6 of the assigned user terminal 1500, the generation AI capable of generating the content element in question is set as the "assigned generation AI," and element-specific generation instruction information is generated for the assigned generation AI.

[0103] As a system that provides virtual experiences, the content provision system 1000 should ideally have the server system 1100 bear all the load related to data generation. However, data generation by the generation AI can sometimes require an enormous amount of computation depending on the generation conditions and required quality settings. When faced with such a case, it could impair the server system 1100's ability to maintain its functionality as a game server.

[0104] In the content provision system 1000 of this embodiment, the generation of content elements with relatively low generation costs can be entrusted to the generation AI of the user terminal 1500. This makes it possible to maintain an appropriate processing load on the server system 1100 even when faced with such cases.

[0105] It is also possible to configure the system without predicting generation costs. For example, depending on the type of content element, it may be predetermined whether to select server-side generation AI3 or element-specific server-side generation AI4 as the responsible generation AI, or terminal-side generation AI5 or element-specific terminal-side generation AI6.

[0106] For example, if the target content is a "fighting game," then content elements generated on the server side may be pre-configured to include "special move animations," "character dialogue," "map destruction," and "ghosts." Incidentally, "map destruction" refers to the phenomenon where background objects of the game stage are destroyed when hit by player characters or attacks unleashed by player characters. A "ghost" is a virtual player character that fights in an average style, for example, based on the past input data of multiple players. Furthermore, content elements generated on the terminal side may include "character accessories (elements that do not affect game progression)," "sound effects (elements that do not affect game progression)," and "visual effects (elements that do not affect game progression)."

[0107] For example, if the target content is a "music game," the "music" itself is set as a content element generated on the server side. The "score (notes)," which is the display that indicates the type and timing of the input, can be set in advance as a content element generated on the terminal side.

[0108] For example, if the target content is a "sports game," then "preset motions" may be set as content elements generated on the server side. "Player avatar," "player avatar accessories," "sound effects," and "visual effects" may be set as content elements generated on the terminal side.

[0109] For example, if the target content is an RPG (Role Playing Game), then "map structure," "background objects," "characters," and "items" can be set as content elements generated on the server side. On the other hand, "object placement on the map," "sound effects," and "visual effects" can be set as content elements generated on the terminal side.

[0110] For example, if the target content is a "board game," then "map" and "characters" can be set as content elements generated on the server side. Conversely, "event details," "sound effects," and "visual effects" can be set as content elements generated on the terminal side.

[0111] Figure 8 shows an example of programs and data stored by the server system 1100. The server system 1100 stores the server program 501, the distribution client program 503, the content initial setup data 510, the trained generation AI model 520, and the content element decomposition pattern data 522 in the IC memory 1152.

[0112] Furthermore, the server system 1100 stores prediction pattern data 540 (see Figure 7), user registration data 600, generated AI registration data 650, server-side generation status information 671 (see Figure 6), received terminal-side generation status information 672 (see Figure 6), and content play information 700.

[0113] Furthermore, the server system 1100 stores the originating generation instruction information 702 (see Figure 2), generation planning data 750, target content data 778, past generation information 780, and the current date and time 900. Of course, other data may also be stored as appropriate.

[0114] The server system 1100 performs the server program 501 on the CPU 1151, thereby realizing the function of a server processing unit 200s, as shown in Figure 9. The server program 501 may include a generation AI program 502 for realizing the functions of server-side generation AI3 and element-specific server-side generation AI4. Alternatively, the generation AI program 502 may be stored separately.

[0115] The server processing unit 200s performs various controls related to content provision. Specifically, the server processing unit 200s includes a user information management unit 202, a payment processing unit 204, an online shopping control unit 206, a server-side generation status acquisition control unit 208, and a terminal information acquisition control unit 210.

[0116] Furthermore, the server processing unit 200s includes a terminal-side generation status information acquisition control unit 212, a content play information acquisition control unit 214, a generation plan unit 216, a terminal-side plan information acquisition control unit 218, a prediction unit 220, a generation planning unit 222, and a generation control unit 224.

[0117] Furthermore, the server processing unit 200s includes a content provision control unit 226, a past generated information storage control unit 228, a server-side generation unit 230, and a timing unit 290.

[0118] The user information management unit 202 executes the prescribed player registration procedure (user registration procedure) and controls the registration and management of user registration data 600 for each user 2.

[0119] The payment processing unit 204 performs the payment processing related to the price of playing the game.

[0120] The online shopping control unit 206 processes online sales of items usable in playing content, such as charging fees and granting purchased items to users.

[0121] The terminal information acquisition control unit 210 controls the acquisition of terminal information from the user terminal 1500.

[0122] The terminal-side generation status information acquisition control unit 212 performs control to acquire terminal-side generation status information indicating the generation status of terminal-side generated AI5 from the user terminal 1500.

[0123] The content play information acquisition control unit 214 controls the acquisition of content play information from the user terminal 1500.

[0124] The generation planning unit 216 performs control to plan generation instructions (cause generation instruction information) that will trigger the generation of content elements.

[0125] The terminal-side intention information acquisition control unit 218 performs control to acquire terminal-side intention information from the user terminal 1500 that indicates an intention to generate an instruction.

[0126] The prediction unit 220 predicts the predicted generation cost related to the generation of content elements.

[0127] The generation planning unit 222 selects whether the generation AI responsible for generating the content elements constituting a given target content should be a server-side generation AI or a terminal-side generation AI. Specifically, the generation planning unit 222 assigns a responsible generation AI to each content element, distributed between the server-side generation AI 3 and the terminal-side generation AI 5, based on the type of content element and the predicted generation cost.

[0128] However, the generation planning unit 222 determines whether or not the content element is an element that is generated using privacy information, and if the determination is affirmative, it selects the terminal-side generation AI as the AI ​​responsible for generating that content element. Furthermore, instead of selecting a responsible generation AI, the generation planning unit 222 searches for replaceable content elements from previously generated content elements and decides to use those content elements as substitutes.

[0129] The generation control unit 224 controls the assigned generation AI to generate the content elements that the assigned generation AI is responsible for. Specifically, the generation control unit 224 creates generation instruction information for generating content elements and provides it to the assigned generation AI.

[0130] The content provision control unit 226 controls the progress of the target content, including content elements generated by the assigned generation AI, and provides the target content to the user terminal 1500. Specifically, the content provision control unit 226 aggregates multiple generated content elements and provides the target content. In this case, if the generation planning unit 222 has decided on a substitute, the content provision control unit 226 uses already generated content elements in place of newly generated content elements to provide the target content.

[0131] The previously generated information storage control unit 228 controls the storage of previously generated content elements as previously generated information 780 (see Figure 8), associating them with the information at the time the content elements were generated.

[0132] The server-side generation unit 230 automatically generates content. The server-side generation unit 230 may also have element-specific server-side generation units 232 for each content element. In the configuration shown in Figure 1, the server-side generation AI3 corresponds to the server-side generation unit 230, and the element-specific server-side generation AI4 (4a, 4b, ...) corresponds to the element-specific server-side generation units 232.

[0133] The timing unit 290 uses the system clock to perform various timings, such as the current date and time (900) and the time spent playing content.

[0134] Returning to Figure 8, the distribution client program 503 is the original client program provided to the user terminal 1500.

[0135] The content initial setup data 510 is prepared separately for each title of content provided by the content provision system 1000 (each title content), and stores various initial setup data for that content. In this embodiment, since the title content is a game, the content initial setup data 510 corresponds to game initial setup data.

[0136] Furthermore, the content initial setup data 510 includes a list of replaceable content elements 512 and event definition data 514.

[0137] The list of replaceable content elements 512 indicates which of the initially configured content elements can be replaced with the desired content. For example, maps, backgrounds, characters, visual effects, sound effects, and items may be set as replaceable content elements (see Figure 5).

[0138] By properly configuring the list of replaceable content elements 512, game developers can maintain the intended game world and game balance.

[0139] The list of replaceable content elements 512 may be a common list covering the entire title content, or it may be set separately for each sub-scenario, game stage, and event included in the title content.

[0140] Event definition data 514 is prepared for each event. Each event definition data 514 stores the event occurrence requirements, a list of the impact of each content element type in that event (the degree of impact on the progress of the title content; a higher value indicates a greater impact), and event execution data.

[0141] By appropriately setting the impact list by content element type, it is possible to achieve both the game creator's intentions regarding the game's progression and the benefits of distributing the AI ​​responsible for generating content between the server system 1100 and the user terminal 1500.

[0142] For example, an event might involve a relatively long and flashy visual effect to impress upon User 2 the appearance of a powerful enemy (for instance, an event where a resurrected enemy boss character destroys a city with powerful magic). If this visual effect is of low quality, it may fail to convey the intended impression to User 2, potentially creating a sense of incongruity in the subsequent development of the game scenario.

[0143] In this case, the influence of visual effects, which are essential for expressing magic, is set to "100" for this event, whereas it is set to "0" for other normal events. Since flashy and relatively long-lasting visual effects are certain to meet the high cost condition for predicted generation costs, we have the server-side generation AI3 handle them from the beginning to obtain high-quality visual effects.

[0144] A pre-trained generative AI model 520 is provided for each element-specific server-side generated AI 4. One pre-trained generative AI model 520 includes the generative AI type and the AI ​​model data.

[0145] The content element decomposition pattern data 522 is prepared for each piece of content containing multiple content elements, from the categories shown as options in the category setting UI 10 (see Figure 3). A single content element decomposition pattern data 522 stores a content category and a list of content element types.

[0146] User registration data 600 is prepared for each user who has completed the prescribed user registration procedure and stores various information (user information) associated with that user.

[0147] One user registration data 600 includes, for example, a user account 601, user information 602, and user-specific past generation history data 610 relating to past generation for that user, as shown in Figure 10. Of course, other information may also be included as appropriate.

[0148] User information 602 includes received terminal information 603 of the user terminal 1500 used by the user, privacy information 604, and a list of available privacy information 605.

[0149] The received terminal information 603 includes information such as CPU type and memory capacity. Privacy information may include user accounts, facial images, age, gender, billing history, subscription contract information, lifestyle patterns, and more. Billing history information includes information such as charges for playing content, charges for purchasing paid items usable in content using online shopping functions, etc., and may also be used as statistical data such as cumulative billing amount and billing frequency.

[0150] Lifestyle pattern information may also be statistical data on the time periods when the user logged in and played content.

[0151] The list of available privacy information 605 indicates the types of information from privacy information 604 that the user has authorized to be used in generating content elements. The available privacy information may also be set by displaying a screen that allows users to select whether or not to allow use when setting / entering privacy information 604 during user registration, and setting the information according to the result of the selection.

[0152] The user-facing past generation history data 610 is created for each intent to generate content and content elements, and stores various data related to the generation. One user-facing past generation history data 610 includes an intent ID 612 and a list of generated content element IDs 614 generated as a result of that intent.

[0153] Returning to Figure 8, the generated AI registration data 650 is created for each generated AI possessed by the server system 1100 and the user terminal 1500. In other words, it is registration data for the generated AI resources available to the server system 1100.

[0154] One generation AI registration data 650 includes, for example, a generation AI ID 652, location information 653, a list of reproducible content element types 654, generation instruction format information 655, generation cost information 656, and generation cost statistics data 657, as shown in Figure 11.

[0155] Location information 653 indicates whether the generated AI corresponds to the server system 1100 or the user terminal 1500. Location information 653 may be set to either a value indicating the server system 1100 or the user terminal ID.

[0156] The list of reproducible content element types 654 is a list of the types of content elements (modals) that the generation AI can generate. In the case of a multimodal AI, multiple types of content elements will be listed.

[0157] The generation instruction format information 655 is information regarding the format and specifications of the generation instruction information (e.g., prompts) for the generation AI.

[0158] The generation cost information 656 stores information regarding the cost of using the generation AI. This information may include, for example, subscription contract information or usage-based billing information. The generation cost information 656 serves as one of the basic data for determining the generation cost related to the generation AI.

[0159] The generation cost statistics data 657 is created for each type of content element generated by the generation AI and stores statistical data on the generation cost. The statistical data may include, for example, generation waiting time statistics, required communication data volume statistics, generation cost statistics, etc.

[0160] Returning to Figure 8, the content play information 700 is prepared for each user and stores various information related to the user's content play. One piece of content play information 700 stores, for example, the player account (the user account of the player user) and game progress information (content progress information).

[0161] The generation planning data 750 includes, for example, as shown in Figure 12, a originating generation intention ID 752 indicating the generation intention that gave rise to the planning, and a content element type list 754 obtained by content element decomposition. It also includes a list of contained privacy information content elements 756 indicating content elements that utilize privacy information, and content element management data 760. Of course, other information may also be included as appropriate.

[0162] The List of Content Elements Containing Privacy Information 756 indicates the types of content elements that have been pre-specified, and the types of content elements for which the use of privacy information is permitted in the Privacy Information Usage Settings UI 12p (see Figure 4).

[0163] One content element management data 760 includes a content element ID 762, a first predicted generation cost 763, a second predicted generation cost 764, a substituted content element ID 765, a responsible generation AI ID 766, and element-specific generation instruction information 770 (see Figure 6).

[0164] Substitutable content element ID 765 indicates a generated content element that can be substituted or replaced for the content element in question.

[0165] The element-specific generation instruction information 770 includes the content element type 771 (the type of information to be generated), generation condition information 772, required quality information 773, supplementary information 774, etc. Supplementary information 774 may include existing content element data of the same type to be used as a base, or existing content element data for referencing features, etc.

[0166] Returning to Figure 8, the target content data 778 is the data of the target content requested by the prospective entity, and includes one or more content element data.

[0167] Past generation information 780 is created for each content element generation and stores various data related to that generation. Each past generation information 780 includes the content element ID, information at the time of generation, and content element data. Of course, other information may also be included as appropriate. "Information at the time of generation" includes the predicted generation cost, element-specific generation instruction information given to the generation AI, etc.

[0168] Figure 13 shows an example of programs and data stored in the user terminal 1500. The user terminal 1500 stores the client program 504, terminal information 808, DL content initial setup data 810, and a trained generation AI model 820 for terminal-side generation AI in its IC memory 1552. The user terminal 1500 also stores terminal-side generation status information 822, game client data 830, UI definition data 840, and the current date and time 900. Of course, other data may be stored as appropriate.

[0169] The user terminal 1500 functions as a terminal processing unit 200t by executing and processing the client program 504 on the CPU 1551. The client program 504 may include a generation AI application program 505 for realizing the functions of terminal-side generated AI 5 and element-specific terminal-side generated AI 6. Alternatively, the generation AI application program 505 may be stored separately.

[0170] The terminal processing unit 200t performs various controls as a game client and various controls to enable the user terminal 1500 to function as a man-machine interface for the content provision system 1000.

[0171] Specifically, as shown in Figure 14, the terminal processing unit 200t includes an operation input information provision control unit 260, a display control unit 262, a terminal information provision control unit 264, a terminal-side generated AI information provision control unit 266, and a terminal-side generation status monitoring unit 268. Furthermore, the terminal processing unit 200t includes a terminal-side generation status information provision control unit 270, a terminal-side intention information provision control unit 272, a content element data provision control unit 274, a terminal-side generation unit 280, and a timing unit 290.

[0172] The operation input information provision control unit 260 controls the transmission of operation input information to the server system 1100. The display control unit 262 performs control to display various images based on data received from the server system 1100.

[0173] The terminal information provision control unit 264 performs control to transmit terminal information 808 (device information) of its own unit in response to a request from the server system 1100.

[0174] The terminal-side generated AI information provision control unit 266 performs control to transmit a list of generated content element types for each terminal-side generated AI 5 and element-specific terminal-side generated AI 6, generation instruction format information, and generation price information in response to a request from the server system 1100.

[0175] The terminal-side generation status monitoring unit 268 monitors and controls the generation status of terminal-side generated AI5 and element-specific terminal-side generated AI6. The monitoring results are saved as terminal-side generation status information 822 (see Figure 13) and are constantly updated with the latest information.

[0176] The terminal-side generation status information provision control unit 270 performs control to send terminal-side generation status information 822 (see Figure 13) in response to a request from the server system 1100.

[0177] The terminal-side intent information provision control unit 272 receives a target content request operation, creates terminal-side intent information 832 (see Figures 2 and 13) related to the target content desired by user 2, and controls its provision to the server system 1100.

[0178] The content element data provision control unit 274 provides the terminal-side generation AI 5 and the element-specific terminal-side generation AI 6 with element-specific generation instruction information received from the server system 1100 to perform content element generation. It then executes control to provide the generated data to the server system 1100.

[0179] The terminal-side generation unit 280 automatically generates content. The terminal-side generation unit 280 may also have element-specific terminal-side generation units 281 for each content element. In the configuration shown in Figure 1, the terminal-side generation AI 5 corresponds to the terminal-side generation unit 280, and the element-specific terminal-side generation AI 6 (6a, 6b, ...) corresponds to the element-specific terminal-side generation units 281.

[0180] Returning to Figure 13, the DL content initial setup data 810 is a copy of the content initial setup data 510 downloaded from the server system 1100.

[0181] The trained generation AI model 820 for terminal-side generation AI stores various data related to terminal-side generation AI 5 and element-specific terminal-side generation AI 6. One trained generation AI model 820 for terminal-side generation AI includes, for example, a generation AI ID, a list of reproducible content element types, generation instruction format information, generation cost information, and the main data of the AI ​​model.

[0182] The game client data 830 stores various data received from the server system 1100 and various data provided to the server system 1100. For example, the game client data 830 includes terminal-side intention information 832 (see Figure 2), received element-specific generation instruction data 834, and terminal-side generated content element data 836 generated on the terminal side. Of course, other information may also be included as appropriate.

[0183] UI definition data 840 is definition data for various UIs (see Figures 3 and 4) used for target content request operations.

[0184] Figures 15 and 16 are flowcharts illustrating the processing flow related to the generation and provision of target content executed by the server system 1100. It is assumed that the player user is registered and logged in.

[0185] The server system 1100 performs the user login process (step S10). In the login process, a predetermined user registration process is performed for new users. At that time, when receiving input of privacy information, the server system accepts the selection of privacy information that can be used to generate content elements.

[0186] Next, the server system 1100 sends a predetermined request to each player's user terminal 1500 to obtain terminal information for each terminal (step S12).

[0187] Next, the server system 1100 searches for a generation AI that can be used to provide the content and creates generation AI registration data 650 (see Figure 11) (step S13). In the search, the generation AI of the server system 1100 and the generation AI of the user terminal 1500 of the player are searched as candidates.

[0188] Furthermore, although user terminals 1500 of users not involved in content play may be included in the search target as terminals that are online with the server system 1100. In this case, the search target may be narrowed down by referring to the lifestyle pattern information in privacy information 604 (see Figure 10).

[0189] Next, the server system 1100 starts control to continuously acquire generation status information for each of the available generation AIs: server-side generation AI 3, element-specific server-side generation AI 4, terminal-side generation AI 5, and element-specific terminal-side generation AI 6 (step S14). From this point onward, the server system 1100 sends a request for terminal-side generation status information 822 (see Figure 13) to the user terminal 1500 at a sufficiently short predetermined interval, and the user terminal 1500 responds to this request.

[0190] Next, the server system 1100 starts providing the title content (step S16). Since the title content in this embodiment is game content, the server system 1100 starts controlling the progress of gameplay. From this point onward, the content play information 700 is updated with the latest play information.

[0191] During gameplay, if the server system 1100 intends to generate new target content due to the occurrence of an event or the like (step S18), it creates cause generation instruction information 702 (step S20). Furthermore, during gameplay, when a predetermined target content request operation is entered on the user terminal 1500, the user terminal 1500 creates terminal-side intention information 832 (see Figure 2) based on the input of that operation. This information is then sent to the server system 1100 (step S18). The server system 1100 retrieves this information from the user terminal 1500 and creates cause generation instruction information 702 (step S20).

[0192] Then, the server system 1100 executes a generation planning process based on the cause generation instruction information 702 (step S22).

[0193] Figures 17 and 18 are flowcharts illustrating the flow of the generation planning process. In the generation planning process, the server system 1100 first initializes the generation planning data 750 (see Figure 12) and decomposes the content elements of the target content (step S40).

[0194] When decomposing elements, content element decomposition pattern data 522 is referenced. If the target content is originally a single content element, the result of the element decomposition will be only the content element that is the target content.

[0195] Furthermore, among the types of content elements obtained through element decomposition, those specified in advance, and those permitted for use in the privacy information usage settings UI12p, are set in the contained privacy information content element list 756.

[0196] Next, the server system 1100 executes Loop A for each content element obtained by decomposition (steps S42 to S122). In loop A, the server system 1100 determines the first predicted generation cost and the second predicted generation cost of the content element to be processed (step S44). If both the first and second predicted generation costs obtained satisfy the predetermined excess conditions (YES in step S46), the server system 1100 searches for replaceable content elements and uses them as substitutes for the content elements to be processed in loop A (step S48).

[0197] Specifically, by referring to the previously generated information 780 (see Figure 8) and the initial content setting data 510, content elements that are at least the same type as the target content and preferably conform to the generation conditions and required quality of the target content are considered replaceable. Then, the corresponding content element is set as replaceable content element ID 765 (see Figure 12), and the data of that content element is copied to the target content data 778 (see Figure 8) as a substitute for the content element to be processed in Loop A. In other words, substitution and replacement are performed, and if it is a content element that has already been generated, it is reused.

[0198] If the predicted cost does not exceed the estimated cost (NO in step S46), the server system 1100 determines whether the only generation AI capable of generating the content elements to be processed is located within the server system 1100. If the determination is positive (YES in step S60), the corresponding generation AI in the server system 1100 is set as the generation AI responsible for the content elements to be processed (step S62), and loop A ends.

[0199] Furthermore, the server system 1100 determines whether the user terminal 1500 is the only generation AI capable of generating the content elements to be processed. If the determination is affirmative (YES in step S64), the server system 1100 sets the generation AI of the corresponding user terminal 1500 as the generation AI responsible for the content elements to be processed (step S66), and ends loop A.

[0200] Moving to Figure 18, if a generation AI capable of generating content elements to be processed exists in both the server system 1100 and the user terminal 1500, the server system 1100 determines whether the content elements to be processed in Loop A are elements that do not affect the progress of the title content (step S70).

[0201] This determination is affirmed, for example, if the influence level of each type of content element (higher values ​​indicate greater influence) does not fall below a predetermined standard. For example, if User 2 is the main developer, the influence levels of the map, background, and characters will be set to "100" because they definitely affect the game's progression. The influence levels of items will be set to "80", visual effects to "0", and sound effects to "0". The predetermined standard value will be "70". In this example, sound effects and visual effects will be determined to be elements that do not affect the game's progression.

[0202] In step S18, if the server system 1100 becomes the initiating entity due to the occurrence of an event, it refers to the settings of the event definition data 514 (see Figure 8) that meets the occurrence requirements to obtain the impact of each type of content element in that event.

[0203] If the element does not affect the progress of the content (YES in step S70), the server system 1100 executes a terminal-side assigned generation AI setting process (details described later) to select and set the assigned generation AI from the user terminal 1500's generation AIs (step S88).

[0204] If the element does not affect the progress of the content (NO in step S70), the server system 1100 determines whether the content element to be processed in loop A is a content element that uses privacy information 604 (step S72). In making this determination, the list of content elements containing privacy information 756 is referred to.

[0205] If step S72 is negative (NO in step S72), the server system 1100 determines whether the first predicted generation cost satisfies the given high cost condition (step S74).

[0206] If step S74 is negative (NO in step S74), the server-side generation AI 3 refers to the originating generation instruction information 702 of the content element to be processed in loop A and determines whether the intentioning entity is the user (step S76). If the determination is positive (YES in step S76), the server system 1100 executes the terminal-side responsible generation AI setting process (details described later) (step S88).

[0207] On the other hand, if step S74 is affirmative (YES in step S74) or if step S76 is negative (NO in step S76), the server system 1100 refers to the server-side generation status information 671 (see Figure 8) to determine whether it falls under a high-load / busy state (step S80). If it does not fall under this state (NO in step S80), the server system 1100 selects a generation AI capable of generating the content element to be processed from the server-side generation AI 3 and the element-specific server-side generation AI 4, and sets it as the responsible generation AI (step S82). If step S80 is affirmative (YES in step S80), the terminal-side responsible generation AI setting process is performed (step S88).

[0208] Figure 19 is a flowchart illustrating the process of setting up the AI ​​generated by the terminal side. In the terminal-side generation AI setting process, the server system 1100 first determines whether there are multiple user terminals 1500 that have generation AIs capable of generating content elements to be processed in Loop A (step S100).

[0209] If there are multiple users (YES in step S100), the server system 1100 calculates priority points for each user terminal 1500 (step S102). Priority points are an index value used to select which of the 1500 relevant user terminals should be prioritized for setting the assigned generation AI. Priority points are calculated as the sum of the second prediction generation point ranking points, the processing capacity ranking points, and the terminal-side generation status ranking points.

[0210] The ranking points for the second prediction generation are set so that the lower the total second prediction generation cost (see Figure 6) for the corresponding user terminal 1500, the higher the score. The processing power ranking points are set so that the higher the processing power estimated from the received terminal information 603 (see Figure 10) of the corresponding user terminal 1500, the higher the score. The terminal-side generation status ranking is set so that the lower the load and the slacker the generation status, as indicated by the received terminal-side generation status information 672 (see Figures 6 and 8) of the corresponding user terminal 1500, the higher the score. Note that the distribution of each bonus point is not limited to the example in Figure 19, and may be set as appropriate.

[0211] Next, the server system 1100 refers to the content play information 700 (see Figure 8) for each of the corresponding user terminals 1500. Then, it deducts a predetermined amount of priority points from the user terminals 1500 whose content progress information indicates a play state that results in a high load / busy state (step S104).

[0212] For example, points will be deducted for user terminal 1500 playing a game stage that requires high-speed rendering of a 3D virtual space containing numerous objects. This includes game stages where players race on a rollercoaster-like course and compete for the best lap time.

[0213] Next, the server system 1100 deducts a predetermined amount of priority points from the user terminal 1500 of high-spending users (step S106). High-spending users are determined by whether the cumulative amount of charges indicated by the billing history information in the privacy information 604 meets the given high-spending conditions.

[0214] Next, the server system 1100 sets the responsible generation AI from the generation AI of the user terminal 1500, which has the highest priority points (step S108).

[0215] If there is only one user terminal 1500 that has a generation AI capable of generating content elements to be processed in Loop A (NO in step S100), the server system 1100 sets the responsible generation AI from among the terminal-side generation AI 5 and element-specific terminal-side generation AI 6 of the corresponding user terminal 1500 (step S110).

[0216] Furthermore, in step S108 or step S110, if there are multiple generation AIs capable of generating the content elements to be processed in loop A, the generation status information 672 on the receiving terminal side (see Figures 8 and 2) may be referred to, and the generation AI with the lightest generation status may be set as the responsible generation AI.

[0217] Alternatively, you could refer to the respective generation cost information 656 (see Figure 11) and prioritize setting the generation AI that is under a subscription agreement as the responsible generation AI. Alternatively, one could refer to the respective generation cost statistics 657 (see Figure 11), calculate the cumulative amount of generation cost from the generation cost statistics, and prioritize assigning the generation AI with the lowest cumulative amount to the assigned generation AI.

[0218] Returning to Figure 18, the server system 1100 then creates element-specific generation instruction information 770 (see Figure 12) for the content elements to be processed in Loop A (step S120), and terminates Loop A (step S122).

[0219] Specifically, the server system 1100 determines the generation conditions, required quality, and supplementary information related to the content elements to be processed from the originating generation instruction information 702 (see Figures 8 and 2). Then, it edits each of these pieces of information according to the generation instruction format information 655 (see Figure 11) of the responsible generation AI to generate element-specific generation instruction information 770.

[0220] Once Loop A has been executed for all the decomposed content elements, we return to Figure 15. Next, the server system 1100 executes loop B for all the decomposed content elements (steps S140 to S164).

[0221] In loop B, the server system 1100 determines whether the content element to be processed in loop B has a specified substitute content element ID 765 (step S142). If it does (YES in step S142), the server system 1100 reads the data of the specified substitute content element as a substitute for the content element to be processed in loop B and sets it in the target content data 778 (step S144).

[0222] If no substitute content element is specified (NO in step S142), proceed to Figure 16, where the server system 1100 determines the location of the responsible generation AI (step S146).

[0223] If the location is the server system 1100 (the server system in step S146), the server system 1100 provides the responsible generation AI with element-specific generation instruction information 770 for the content elements to be processed in Loop B, which were created in the generation planning process. Then, it causes the content elements to be generated (step S148). The data of the generated content elements is stored and aggregated in the target content data 778.

[0224] Next, the server system 1100 creates and saves past generation information 780 (see Figure 8) of the generated content elements (step S160), and terminates loop B (step S164).

[0225] On the other hand, if the location is user terminal 1500 (the user terminal in step S146), the server system 1100 sends element-specific generation instruction information 770 along with the assigned generation AIID to the corresponding user terminal 1500 (step S150).

[0226] The user terminal 1500, which is the device located there, stores this as received element-specific generation instruction data 834 (see Figure 13), and provides the received element-specific generation instruction information 770 to the assigned generation AI, causing it to generate content elements. The user terminal 1500, which is the device located there, sends back the terminal-generated content element data 836 (see Figure 13) of the generated content element to the server system 1100.

[0227] The server system 1100 acquires the terminal-generated content element data 836, stores it in the target content data 778, and aggregates it (step S152).

[0228] Next, the server system 1100 creates and saves past generation information 780 of the generated content elements (step S160), updates the generation cost statistics data 657 (see Figure 11) (step S162), and terminates loop B (step S164).

[0229] Once Loop B has been executed for all content elements obtained through element decomposition, the server system 1100 updates the user-facing past generation history data 610 (see Figure 10) (step S170).

[0230] Then, the server system 1100 starts providing the target content (step S172). At this time, if the target content is replaceable content (for example, a character) that is initially set in the title content's content initial setting data 510, the server system 1100 controls the replacement of the replaceable content with the target content.

[0231] The server system 1100 returns to step S18 until a given termination condition relating to the title content is met (NO in step S174). When the given termination condition related to the title content is met (YES in step S174), the server system 1100 terminates the series of processes.

[0232] As described above, this embodiment makes it possible to reduce the processing load on the server system that provides content including content elements generated by the generation AI to the user. In other words, the generation of content elements constituting the target content is distributed between the generation AI of the server system 1100 and the generation AI of the user terminal 1500, making it possible to reduce the processing load on the server system 1100 related to generation.

[0233] [Variation] Although examples of embodiments to which the present invention is applied have been described above, the forms to which the present invention can be applied are not limited to the above forms, and components can be added, omitted, or modified as appropriate.

[0234] (Variation 1) For example, the content provision system 1000 may be implemented using a P2P (Peer to Peer) architecture with multiple user terminals 1500. In this case, programs and data corresponding to the functional division are stored in the user terminals 1500, and the functions corresponding to the server processing unit 200s in the above embodiment are distributed and implemented by the user terminals 1500, which act as P2P nodes. The same effects as in the above embodiment can be obtained with this configuration as well.

[0235] (Variation 2) In the above embodiment, the server-side generated AI3 and the element-specific server-side generated AI4 were described as being installed on the server system 1100, but the invention is not limited to this. The server-side generated AI3 and the element-specific server-side generated AI4 may also be external devices that can be connected via the network 9. Similarly, the terminal-side generated AI5 and the element-specific terminal-side generated AI6 were described as being installed on the user terminal 1500, but the invention is not limited to this. The terminal-side generated AI5 and the element-specific terminal-side generated AI6 may also be external devices that can be connected via the network 9.

[0236] (Variation 3) Furthermore, the input for the target content request operation on the user terminal 1500 may be made available in natural language, allowing the user to instruct in their own words what kind of content elements to generate.

[0237] Specifically, the server system 1100 is equipped with a deeply trained AI model capable of generating cause-generating instruction information 702 from input natural language (see Figure 2). Instead of displaying the category setting UI10 or the detailed setting UI12 (see Figure 3), the user terminal 1500 displays a request setting screen W10 that allows input using natural language text, for example, as shown in Figure 20. The text entered on this setting screen may then be used in place of the terminal-side intention information 832 (see Figure 2).

[0238] (Variation 4) In the above embodiment, an example was given in which the target content is generated and provided while the title content is being provided. However, this embodiment may also be applied in cases where content equivalent to the title content is generated as the target content.

[0239] (Variation 5) In the above embodiment, the main control of content provision was performed by the server system 1100, but it is also possible to configure the system so that the user terminal 1500 performs this control. In this configuration, it can be said that the roles of the server system 1100 and the user terminal 1500 in the above embodiment are reversed.

[0240] Figure 21 shows an example of the programs and data stored in the user terminal 1500B in this configuration. The user terminal 1500B stores the client program 504B, terminal information 806, DL content initial setup data 810, trained generation AI model 820 for terminal-side generation AI, terminal-side generation status information 822, and received server-side generation status information 824.

[0241] Furthermore, the user terminal 1500B stores terminal-side intention information 832, UI definition data 840, content element decomposition pattern data 522, prediction pattern data 540, generated AI registration data 650, and content play information 700.

[0242] Furthermore, it stores the cause generation instruction information 702, generation planning data 750, target content data 778, and past generation information 780. The content element decomposition pattern data 522 and the prediction pattern data 540 are downloaded and stored from the server system 1100B as needed, similar to the client program 504B.

[0243] The user terminal 1500B executes the client program 504B to realize the function of the terminal processing unit 200t shown in Figure 22. The terminal processing unit 200t in this configuration includes a user information management unit 202, a terminal-side generation status monitoring unit 268, a server-side generation status information acquisition control unit 269, and a generation planning unit 216. It also includes a planning operation reception control unit 219, a prediction unit 220, a generation planning unit 222, a generation control unit 224, a content progress control unit 227, a past generation information storage control unit 228, a terminal-side generation unit 280, and a timing unit 290.

[0244] The server-side generation status information acquisition control unit 269 acquires server-side generation status information 671, which indicates the generation status of server-side generated AI3, from the server system 1100B at a given timing (for example, a predetermined period) and stores it as received server-side generation status information 824.

[0245] The intent operation reception control unit 219 performs control related to the reception of intent operations by user 2. Specifically, it performs control related to the reception of user 2's operation to select the category of the target content in response to the target content request operation, and sets the terminal-side intent information 832.

[0246] The content progress control unit 227 controls the progress of the target content, which includes content elements generated by the AI ​​responsible for generating content.

[0247] Figure 23 shows examples of programs and data stored by the server system 1100B. The server system 1100B stores the server program 501B, the distribution client program 503B, the content initial setup data 510, and the trained generation AI models 520 for each element of the server-side generation AI 4. It also stores content element decomposition pattern data 522, prediction pattern data 540, user registration data 600, server-side generation status information 671, and the current date and time 900.

[0248] The server system 1100B implements the functions of a server processing unit 200s as shown in Figure 24 by executing the server program 501B. The server processing unit 200s in this configuration includes a payment processing unit 204, an online shopping control unit 206, and a server-side generated AI information provision control unit 282. It also includes a server-side generation status monitoring unit 284, a server-side generation status information provision control unit 286, a content element data provision control unit 288, and a timing unit 290.

[0249] The server-side generated AI information provision control unit 282 performs control to transmit a list of generated content element types for each server-side generated AI3 and element-specific server-side generated AI4, generation instruction format information, and generation price information in response to a request from the user terminal 1500B.

[0250] The server-side generation status monitoring unit 284 monitors and controls the generation status of server-side generated AI3 and element-specific server-side generated AI4. The monitoring results are saved as server-side generation status information 671 and are constantly updated with the latest information.

[0251] The content element data providing control unit 288 gives the element-by-element generation instruction information provided from the user terminal 1500B to the server-side generation AI 3 and the element-by-element server-side generation AI 4 to cause the generation of content elements. Then, it executes control to provide the generated data to the user terminal 1500B.

[0252] In this configuration, the user terminal 1500B executes the content executed by the server system 1100 in the processing flow described in FIGS. 15 to 19, and the server system 1100B executes the content executed by the user terminal 1500 in the processing flow described in FIGS. 15 to 19.

[0253] According to this configuration, the same effects as those of the above-described embodiment can be obtained.

Explanation of Signs

[0254] 3... Server-side generation AI 4... Element-by-element server-side generation AI 5... Terminal-side generation AI 6... Element-by-element terminal-side generation AI 200s... Server processing unit 200t... Terminal processing unit 208... Server-side generation status acquisition control unit 210... Terminal information acquisition control unit 212... Terminal-side generation status information acquisition control unit 214... Content play information acquisition control unit 216... Generation intention unit 218... Terminal-side intention information acquisition control unit [[ID=…]] 220... Prediction unit 222... Generation planning unit 224... Generation control unit 226... Content providing control unit 228... Past generation information storage control unit 230... Server-side generation unit 264... Terminal information providing control unit 270... Terminal-side generation status information providing control unit 272... Terminal-side intention information providing control unit 280... Terminal-side generation unit 501… Server program 512… Replaceable content element list 520… Trained generation AI model 522… Content element decomposition pattern data 540… Prediction pattern data 560… Prediction generation cost 562… Prediction load information 600… User registration data 602… User information 603… Terminal information 604… Privacy information 650… Generation AI registration data 657… Generation cost statistical data 671… Server-side generation status information 672… Terminal-side generation status information 700… Content play information 702… Origin generation instruction information 750… Generation planning data 760… Content element management data 763… First prediction generation cost 764… Second prediction generation cost 765… Replaceable content element ID 766… Responsible generation AI ID 770… Element-by-element generation instruction information 778… Target content data 780… Past generation information 820… Trained generation AI model for terminal-side generation AI 822… Terminal-side generation status information 832… Terminal-side intention information 1000… Content providing system 1100… Server system 1500… User terminal

Claims

1. A server system that controls the progress of content and provides said content to the user terminal, The user terminal has the function of a terminal-side generation unit, A generation planning means for selecting whether the generation unit responsible for generating content elements constituting a given target content is a server-side generation unit or a terminal-side generation unit, A generation control means that controls the assigned generation unit to generate the content elements that the assigned generation unit is responsible for, Content provision control means that controls the progress of the target content, which includes the content elements generated by the responsible generation unit, and provides the target content to the user terminal. A server system equipped with the following features.

2. There are multiple content elements as described above. The generation planning means distributes the content elements between the server-side generation unit and the terminal-side generation unit, assigning the respective generation units to each content element. The server system according to claim 1.

3. The content provision control means performs control to aggregate the generated multiple content elements and provide the target content. The server system according to claim 2.

4. The generation planning means selects the responsible generation unit based on the content elements. The server system according to claim 1.

5. The generation planning means selects the responsible generation unit based on the degree to which the content element influences the progress of the target content. The server system according to claim 4.

6. Predictive means for predicting the generation cost related to the generation of the content elements, Furthermore, The generation planning means selects the responsible generation unit based on the predicted generation cost. The server system according to claim 1.

7. The aforementioned predicted generation cost includes predicted load information related to the generation of the content elements, The generation planning means selects the responsible generation unit based on the predicted load information. The server system according to claim 6.

8. The generation planning means selects the responsible generation unit based on user information. The server system according to claim 1.

9. The aforementioned user information includes privacy information, The generation planning means determines whether the content element is an element generated using the privacy information, and if the determination is affirmative, it selects the terminal-side generation unit as the responsible generation unit for the content element. The server system according to claim 8.

10. Terminal information acquisition control means for acquiring terminal information from the user terminal, Furthermore, The generation planning means selects the responsible generation unit based on the terminal information. The server system according to claim 1.

11. Content play information acquisition control means for acquiring content play information related to the user terminal, Furthermore, The generation planning means selects the responsible generation unit based on the content play information. The server system according to claim 1.

12. A terminal-side generation status information acquisition control means that performs control to acquire terminal-side generation status information indicating the generation status of the terminal-side generation unit from the user terminal, Furthermore, The generation planning means selects the responsible generation unit based at least on the terminal-side generation status information. The server system according to claim 1.

13. A generation planning means that controls the generation instructions that trigger the generation of the aforementioned content elements, A terminal-side intention information acquisition control means for acquiring terminal-side intention information that intends to give the generation instruction from the user terminal, Furthermore, The generation planning means selects the responsible generation unit based on whether the generation instruction is from the generation intention means or based on the terminal-side intention information. The server system according to claim 1.

14. The content element includes replaceable content elements that can be replaced during the progress control of the content. The generation planning means selects the terminal-side generation unit as the responsible generation unit for the replaceable content element, The content provision control means performs control to replace the content element using the replaceable content element during the content progression control. The server system according to claim 1.

15. The generation control means creates generation instruction information for generating the content elements and provides it to the responsible generation unit, and performs control accordingly. The server system according to claim 1.

16. A past generation information storage means that stores the generated content elements as past generation information, associated with the information at the time of generation. The server system according to claim 1, further comprising:

17. The generated information includes at least one of the following: information on the type and / or quality of the content elements, and information on the generation cost. The generation planning means, based on the information generated, decides to substitute the previously generated content elements included in the past generation information for the selection of the responsible generation unit. The content provision control means, when the substitution is determined by the generation planning means, controls the provision of the target content using the generated content elements included in the past generation information. The server system according to claim 16.

18. A program that allows a user terminal to communicate with a server system and control the progress of content, A generation planning means for selecting whether the generation unit responsible for generating content elements constituting a given objective content is the server-side generation unit of the server system or the terminal-side generation unit of the user terminal, A generation control means that controls the assigned generation unit to generate the content elements that the assigned generation unit is responsible for generating, Content progress control means for controlling the progress of the target content, which includes the content elements generated by the designated generation unit. A program to enable the aforementioned user terminal to function.

Images