Behavior change induction system and behavior change induction method

Abstract

To design an intervention measure for guiding an action of an opponent to be changed into an action (target action) preferable for the opponent.SOLUTION: A biological change inducing system (300) includes means (131A) for reproducing a thinking pattern of a target subject using a prediction model machine-learned using biological information of the target subject as teacher data, means (131B) for predicting a reaction of the target subject to a current external stimulus to the target subject on the basis of the thinking pattern of the target subject using the current external stimulus as an input, and means (131C) for designing an interventional measure for inducing an action of the target subject to a target action.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to a behavior change induction system, a behavior change induction method, and a program for causing a computer to execute the method, wherein the system machine-learns the thought patterns and behavior styles of a target person based on the target person's past reactions (verbal and behavioral actions) taken when receiving external stimuli, predicts the reactions (verbal and behavioral actions) that the target person will take when receiving current external stimuli or external stimuli expected in the near future, and designs intervention measures for inducing the target person to change their behavior to a target behavior (preferred behavior) when the predicted reaction is not preferable. [Background Art]

[0002] The United States Department of Defense (the Pentagon) published a document dated January 7, 2010 titled "Psychological Operations" (Non-Patent Document 1). Within the scope of the concept of "Psychological Operations" (hereinafter referred to as "PsyOp") defined in this document, conventionally, the diffusion of disinformation by "State Players" (state organs) sponsored by militaries and governments has been recognized as the core of PsyOp or as one of the new tactical weapons. By diffusing disinformation, it is possible to cause the government and military of an opposing country to make wrong judgments and take wrong actions, and support physical attacks such as missile strikes by one's own country. Thus, the diffusion of disinformation is positioned as a new tactical weapon replacing physical attacks.

[0003] In modern times that have moved from the era of PsyOp to the era of cognitive warfare, information (including disinformation) is a strategic weapon aimed at changing the cognition of each individual citizen, including the upper echelons of an enemy country, to align with the goals of one's own country, and has come to be used for attacking the cognition of each individual for purposes such as influencing civilian opinion in enemy countries and shaping world public opinion. While PsyOp primarily concerns the dissemination of disinformation as a tactical weapon in existing war domains such as land, sea, and air, in modern cognitive warfare, the cognitive domain itself has become the war domain. It is necessary to understand that, in addition to tactical aspects such as disrupting or destroying enemy leadership, it also functions as a strategic weapon targeting the cognitive transformation of individual civilians. This is similar to how cyber technology, once an offensive technology supporting physical tactical weapons in traditional war domains, has now become the war domain itself.

[0004] In modern cognitive warfare, the information generated and used goes beyond simply disrupting the enemy's command structure; it is integrated and produced based on some long-term benefit of one's own country, and in many cases, a large amount of information generated by AI, such as generative AI, is highly combined and disseminated. Furthermore, in modern times, the information used includes not only linguistic and other symbolic information, but also quasi-symbolic information such as subliminal messages. Moreover, non-symbolic brain information processing such as biofeedback is becoming possible, allowing for direct intervention in the target person's thoughts and rewriting of information. To utilize these technologies, or to defend against such attacks, it is necessary to abstract all patterns from a large amount of information at different levels of abstraction, such as from the five senses and language, recognize the enemy's intentions, and utilize that information. To achieve this, it is necessary to operate multiple AIs in a cooperative and distributed manner and perform inference and analysis at all levels of abstraction. [Prior art documents] [Non-patent literature]

[0005] [Non-Patent Document 1] https: / / irp.fas.org / doddir / dod / jp3-13-2.pdf [Overview of the Initiative] [Problems that the invention aims to solve]

[0006] As mentioned above, in the modern era, information, regardless of whether it is true or false, has become a strategic weapon aimed at altering the perceptions of every citizen, including the top brass of an enemy country, to align with one's own country's objectives. Furthermore, handling information in a way that aligns with one's objectives is an extremely important strategy, not only in the field of strategic weapons but also in the field of business. The same can be said for individuals in their normal daily lives. However, in the fields of strategy and business, methods for utilizing such strategically aligned information are still not sufficiently developed. This invention has been made in view of the current situation and aims to provide a system and method for utilizing information in accordance with the above-mentioned strategy. Furthermore, since there is no prior art similar to the present invention, the above-mentioned document concerning PsyOp published by the Pentagon is cited as prior art. [Means for solving the problem]

[0007] To achieve this objective, the present invention, in a first aspect, is a behavior change induction system (300) for designing intervention measures to induce behavior change in a target person (172), comprising: a storage means (132) for storing biometric information including the target person's (172) responses to past external stimuli to the target person (172); a thought pattern reproduction means (131A) for reproducing the target person's (172) thought patterns or behavior patterns using a predictive model (210) trained with the biometric information as training data (200); and a system for storing current external stimuli to the target person (172) The present invention provides a behavior change induction system (300) comprising: a response estimation means (131B) that takes expected future external stimuli as input and outputs the expected response of the target subject (172) to those external stimuli based on the thought patterns or behavior patterns of the target subject (172) reproduced by the thought pattern reproduction means (131A); and an intervention design means (131C) that designs an intervention to guide the behavior of the target subject (172) toward a target behavior in accordance with the expected response of the target subject (172) output by the response estimation means (131B).

[0008] Preferably, the aforementioned biological information includes brain pattern information of the target subject (172). For example, the intervention design means (131C) preferably designs at least one of the following: an intervention to guide the target person (172) toward weakening the emotion caused by the external stimulus, and an intervention to guide the target person (172) toward replacing the emotion caused by the external stimulus with another emotion. In a second aspect, the present invention provides a portable wireless communication device (100) incorporating the above-described behavioral change induction system (300).

[0009] In a third aspect, the present invention relates to a behavior change promotion system for designing and implementing intervention measures to induce behavioral change in a target person (172), wherein the behavior change promotion system comprises a behavior change induction system (300) and a portable wireless communication device (173) owned by the target person (172), the behavior change induction system (300) includes a storage means (132) for storing biological information including the target person's (172) responses to past external stimuli to the target person (172), a thought pattern reproduction means (131A) for reproducing the target person's (172) thought patterns or behavioral patterns using a predictive model (210) trained with the biological information as training data (200), and the thought pattern of the target person (172) reproduced by the thought pattern reproduction means (131A) with current or expected future external stimuli to the target person (172) as input. The behavior change guidance system (100) provides a behavior change promotion system comprising: a response estimation means (131B) that outputs the expected response of the target person (172) to external stimuli based on their behavior or behavioral patterns; and an intervention plan design means (131C) that designs an intervention plan for guiding the target person (172)'s behavior toward a target behavior in accordance with the expected response of the target person (172) output by the response estimation means (131B); the portable wireless communication device (173) has a built-in application (174) for designing an intervention plan based on the intervention plan; the behavior change guidance system (100) transmits the intervention plan to the portable wireless communication device (173) of the target person (172); and the application (174) of the portable wireless communication device (173) that receives the intervention plan formulates an intervention behavior customized for each target person (172) based on the intervention plan.

[0010] As a fourth aspect of this invention, the present invention provides a method for inducing behavioral change, which involves designing an intervention measure to induce behavioral change in a target individual (172). A program to make a computer run and a first that stores biological information including the target subject's (172) response to past external stimuli. process (S110) and a second predictive model (210) which has been machine-trained using the biometric information as training data (200) to reproduce the thought patterns or behavioral patterns of the target person (172) process (S130) and a third process (S140) which takes current external stimuli or expected future external stimuli to the target person (172) as input and outputs the expected response of the target person (172) to those external stimuli based on the thought patterns or behavioral patterns of the target person (172) reproduced in the second process (S130), and a fourth process (S160) which designs intervention measures to guide the behavior of the target person (172) toward a target behavior according to the expected response of the target person (172) output in the third process (S140), A program that makes a computer execute To provide.

[0011] The fourth of the above process (S160) Design at least one of the following: an intervention to weaken the emotion that the target person (172) has experienced due to the external stimulus, and an intervention to replace the emotion that the target person (172) has experienced due to the external stimulus with another emotion. Being It is preferable. As a fifth aspect of the present invention, a method for inducing behavioral change, which involves designing an intervention measure to induce behavioral change in a target person (172). A program to make a computer run and a first that stores biological information including the target subject's (172) response to past external stimuli. process (S110) and a second predictive model (210) which has been machine-trained using the biometric information as training data (200) to reproduce the thought patterns or behavioral patterns of the target person (172) process (S130) and the current external stimulus or expected future external stimulus to the target person (172) are inputs, and the second processBased on the thought patterns or behavioral patterns of the target subject (172) reproduced in (S130), a third system outputs the expected responses of the target subject (172) to those external stimuli. process (S140) and the third process In (S140), in response to the expected response of the target subject (172) output, a fourth intervention policy is designed for intervention measures to guide the behavior of the target subject (172) toward the target behavior. process (S160) and a fifth application built into the portable wireless communication device owned by the target person (172) that formulates a customized intervention action for each target person (172) based on the intervention policy. process (S170) and, A program that makes a computer execute To provide.

[0012] The fourth of the above process (S160) Design at least one of the following: an intervention to weaken the emotion that the target person (172) has experienced due to the external stimulus, and an intervention to replace the emotion that the target person (172) has experienced due to the external stimulus with another emotion. Being Preferred 。 The symbols in parentheses are included to indicate the correspondence with the embodiments described later, and are not intended to limit the scope of the claims.

[0013] As described above, the present invention aims to achieve behavioral change in target individuals. For example, one application of this invention in the military field could be to make the leader of an enemy country feel that their country has the upper hand, thereby causing them to decide to avoid war against their country, or to cause the enemy leader to unconsciously choose actions that give their country the upper hand, leading them to feel that their operation has backfired. The field of application of the present invention is not limited to the military field, and can also be applied to the business field. For example, by implementing the present invention in the marketing field, consumers can be encouraged to purchase the company's own products. Existing behavior change approaches (methods for changing the behavior of others so that others take the behavior desired by oneself), particularly behaviorist approaches, have mainly adopted an "event-driven type model" that predicts future events based on past events. Since this event-driven approach involves the process of inferring an output event that is correlated with a specific input event, time-series information of events is essential, and the following fundamental limitations exist.

[0014] The event-driven model only reacts based on input events. That is, the event-driven model is inherently reactive. For this reason, it can only handle known patterns and predictable events at all times, and it is impossible to take pre-emptive action. For example, in military operations, this results in a reactive posture toward the enemy, which inevitably places one at a constant disadvantage. The event-driven model depends only on past data. For this reason, when unknown or unexpected events occur, it cannot respond appropriately and lacks flexibility. For example, for historically unprecedented events such as a pandemic of a novel virus or the emergence of an unknown technology, it is impossible to formulate countermeasures using only past data. This problem is particularly prominent in the military and security fields, where the ability to immediately respond to unpredictable new threats is required. Furthermore, conventional behaviorist approaches have abstracted people's behavioral tendencies to a certain extent and handled them as a whole, thus failing to address the individual behavioral tendencies of each person.

[0015] In general inference systems such as large language models (LLMs), a wide range of text, audio, image, and video information is used as input data, rather than targeting a specific person. This data often consists of fictional works or descriptions of emotions and does not necessarily accurately reflect the thought patterns of actual humans. Therefore, inference systems built on this general data may not accurately represent the thought and behavior patterns of a specific individual. To solve these problems, the present invention employs an approach that does not rely on time-series event data, but is based on the direct analysis of the target subject's biological information (e.g., brain patterns obtained by a brain imaging device).

[0016] In this approach, as will be described later, three different inference systems (thought reproduction AI, behavior prediction AI, and behavior change AI) are used. Each of these analyzes the individual biometric information (such as brain patterns) of the target person, predicts the target person's future behavior, and designs specific intervention measures for those behaviors. Furthermore, as described later, in this invention, multiple identical inference systems are operated in parallel to reproduce the target person's cognitive model in more detail. This parallelization allows for the reproduction of multiple thought patterns inherent in the target person, enabling optimal inference tailored to the situation. It also allows for real-time prediction of the target person's potential for behavioral change and the rapid design of effective intervention measures. Furthermore, by using a simulation-based behavior prediction system to pre-simulate the behavioral change intervention measures proposed by the above inference system, and by repeatedly verifying and evaluating them, the optimal intervention measures can be created.

[0017] This system is particularly valuable in military domains where rapid decision-making is crucial and the battle situation can change instantaneously. It enables the prediction of enemy attack intentions in advance and the planning and execution of swift defensive actions. As described above, the present invention is based on an approach that directly analyzes the biological information of target individuals. Specifically, the present invention breaks away from the conventional constraints of relying on time-series event data and adopts a cognitive model architecture based on cognitive science to achieve the objective of "who to encourage, when, and what kind of action to encourage." [Effects of the Invention]

[0018] According to the present invention, it is possible to analyze the target person's brain patterns and other biological information in real time, predict the target person's behavior before an event occurs, and guide the target person to change their behavior. This allows for strategic countermeasures to be taken in advance of the target person, overcoming the conventional reactive, after-the-fact approach. In this invention, behavioral prediction is performed directly from current biological information, without depending on past event patterns. Therefore, it is possible to quickly adapt to new events and changes and provide countermeasures. In this invention, detailed biological information of each target individual is directly analyzed, making it possible to predict the individual behavior of each target individual. This allows for a more accurate understanding of each target individual's actual cognitive patterns and behavioral tendencies, and enables the creation of more precisely individualized intervention strategies for each target individual.

[0019] According to the present invention, it is possible to generate customized intervention behaviors for each target individual, thereby improving the effectiveness of the intervention for each target individual. Furthermore, in the process leading to intervention actions, it is possible to reduce the use of computing resources by sharing roles between cloud computers and edge computers. This allows large-scale campaigns and information dissemination activities to be carried out efficiently with limited resources. [Brief explanation of the drawing]

[0020] [Figure 1]This is a block diagram showing an example of the structure of a behavioral change induction system according to the first embodiment of the present invention. [Figure 2] This is a conceptual diagram showing the structure of the first program. [Figure 3] This is a conceptual diagram illustrating the functions of the first to third programs. [Figure 4] Figure 1 is a flowchart showing the operation of the behavioral change induction system. [Figure 5] This is a conceptual diagram illustrating the collaborative relationships between the first, second, and third programs. [Figure 6] This is a conceptual diagram of the construction and maintenance of a cognitive model by the first program (thought-reproducing AI). [Figure 7] This figure shows an example of biometric information of a target individual. [Figure 8] This figure shows an example of behavioral information of target individuals. [Figure 9] This figure shows examples of stimulus information that target individuals receive from their external environment. [Figure 10] This is a conceptual diagram of the behavior prediction system, which is composed of the second program.

[0021] [Figure 11] This is a conceptual diagram of a behavioral change system composed of the third program. [Figure 12] This is an overview diagram of the intervention delivery system. [Figure 13] This figure shows an example of inputs and outputs (intervention policy signals and intervention actions) in the security domain. [Figure 14] This figure shows an example of input and output (intervention policy signals and intervention behaviors) in the field of marketing. [Figure 15] This figure shows an example of input and output (intervention policy signal and intervention behavior) in the entertainment domain. [Modes for carrying out the invention]

[0022] Figure 1 is a block diagram showing an example of the structure of a portable wireless communication device 100 equipped with a behavioral change induction system according to the first embodiment of the present invention. The portable wireless communication device 100 comprises a communication unit 110, a control unit 120, an external memory (hard disk) 130, an input / output unit 140, and an antenna 150. Of these, the behavioral change induction system 300 consists of a control unit 120 and an external memory (hard disk) 130. The communication unit 110 is connected to the antenna 150 and transmits and receives data wirelessly with other wireless communication devices via the antenna 150. The communication unit 110 includes a wireless receiving unit 111, a wireless transmitting unit 112, and a changeover switch 113.

[0023] The wireless receiver 111 demodulates data received from other wireless communication devices and sends it to the control unit 120. The wireless transmitter 112 modulates the data output from the control unit 120 and transmits it to other mobile phones or other wireless communication devices via the antenna 150. The changeover switch 113 receives a signal from the control unit 120 and switches between transmitting (modulation) and receiving (demodulation) according to that signal. The control unit 120 consists of a central processing unit (CPU) 121, a first memory 122 consisting of ROM, a second memory 123 consisting of RAM, an input interface 124 for transferring various commands and data input to the control unit 120 to the central processing unit 121, an output interface 125 for outputting processing results executed by the central processing unit 121 to the outside, and a bus 126 connecting the central processing unit 121 to the first memory 122, the second memory 123, the input interface 124, and the output interface 125, respectively.

[0024] The first memory 122 stores various control programs executed by the central processing unit 121 and other data that cannot be rewritten. The second memory 123 stores various data and parameters and also provides an operating area for the central processing unit 121. That is, data or programs temporarily required by the central processing unit 121 to execute various control programs are read from the external memory 130 and temporarily stored in the second memory 123. The central processing unit 121 controls the overall operation of the portable wireless communication device 100 together with the OS (Operating System). Specifically, it reads the first to third programs 131A, 131B, and 131C (described later) from the external memory 130 and executes those programs. In other words, the central processing unit 121 operates according to the first to third programs 131A, 131B, and 131C stored in the external memory 130. As will be described later, the central processing unit 121 outputs a predetermined result for a predetermined input via a trained evaluation model.

[0025] The input / output unit 140 consists of an operation unit 141, a display 142, a speaker 143, and a microphone 144. The operation unit 141 consists of, for example, a numeric keypad, and various data are input to the control unit 120 via the operation unit 141. The display 142 is, for example, a liquid crystal display and displays the results of calculations performed by the control unit 120 and other data on the screen. Audio data received from an external source is output from the speaker 143 via the wireless receiver 111 of the communication unit 110. The voice of the user of the portable wireless communication device 100 is collected via the microphone 144 and output to the outside via the wireless transmitter 112 of the communication unit 110. The external memory (hard disk) 130 consists of an application section 131 and a data storage section 132.

[0026] The data storage unit 132 consists of a biometric information storage unit 132A, which stores the brain patterns and other biometric information of the target subject, as described later, and a sub-data storage unit 132B, which stores various data about the target subject other than biometric information. The application unit 131 stores the OS 131S, which controls the overall operation of the portable wireless communication device 100, as well as the first program 131A, the second program 131B, and the third program 131C. The first program 131A, the second program 131B, and the third program 131C are programs that operate as a thought reproduction system, an action prediction system, and an action modification system, respectively, as described later. Figure 2 is a conceptual diagram showing the configuration of the first program 131A. As shown in Figure 2, the first program 131A consists of a training data input program 135A, an evaluation model generation program 135B, and an evaluation result output program 135C.

[0027] The second program 131B and the third program 131C, like the first program 131A, are composed of these three programs. The training data input program 135A is a program for inputting training data into the central processing unit 121 in order to generate an evaluation model using machine learning. The evaluation model generation program 135B generates each evaluation model using machine learning with the training data input by the training data input program 135A. An algorithm used to generate an evaluation model is, for example, a neural network. The evaluation result output program 135C outputs the trained evaluation model generated by the evaluation model generation program 135B to the central processing unit 121.

[0028] Figure 3 is a conceptual diagram illustrating the functions of the first to third programs 131A, 131B, and 131C. As shown in Figure 3, the training data 200 is sent to the evaluation model generation program 135B by the training data input program 135A, and the evaluation model generation program 135B generates an evaluation model 210 based on the training data 200. This evaluation model 210 undergoes continuous machine learning and becomes a trained evaluation model. The trained evaluation model 210 is input with data 220 corresponding to the first to third programs 131A, 131B, and 131C. For example, the first program 131A is input with the biological information of the target person sent from the biological information storage unit 132A and other data about the target person sent from the sub-data storage unit 132B as data 220. The trained evaluation model 210 generates predetermined evaluation results (for example, the target person's thinking patterns, the target person's behavior predictions, intervention measures, etc., as described later) based on the data 220, and outputs these results as output 230.

[0029] The evaluation results generated by the trained evaluation model 210 are output to the central processing unit 121 by the evaluation result output program 135C, and are displayed on the display 142 or printed via a printer (not shown) as needed. Figure 4 is a flowchart showing the operation of the portable wireless communication device 100. The following outlines the various operations performed by the portable wireless communication device 100, referring to Figure 4. First, the target person's thought patterns, behavioral patterns, and other biometric information, as well as various other information about the target person, are collected from the target person's past responses (speeches and actions) when the target person received external stimuli. The biometric information is stored in the biometric information storage unit 132A of the external memory 130, and the other information is stored in the sub-data storage unit 132B (step S110). Biological information also includes, for example, brain pattern information obtained by brain imaging devices (e.g., fMRI (functional Magnetic Resonance Imaging), EEG (Electroencephalogram), etc.).

[0030] These biometric and other information are sent as training data to the evaluation model generation program 135B by the training data input program 135A of the first program 131A, and the evaluation model generation program 135B generates an evaluation model 210 using machine learning (step S120). Furthermore, the evaluation model generation program 135B uses a machine learning-based evaluation model to reproduce the thought patterns and behavioral patterns of the target individual (step S130). The reproduced thought patterns and behavioral patterns of the target subject are sent to the second program 131B by the evaluation result output program 135C. In the second program 131B, current external stimuli or external stimuli expected to occur in the near future to the target subject are used as input to the evaluation model, and the responses (speeches and actions) that the target subject will take when receiving those external stimuli are predicted via the evaluation model already generated by the evaluation model generation program 135B (step S140).

[0031] The response that the target person is likely to take is sent to the third program 131C via the evaluation result output program 135C of the second program 131B, and the third program 131C determines whether the target person's response is favorable to the user of the portable wireless communication device 100 (step S150). If the target person's response is favorable (YES in step S150), no intervention measures for the target person (described later) are considered, and the operation of the portable wireless communication device 100 ends here. If the target person's response is unfavorable (NO in step S150), the evaluation model generation program 135B of the third program 131C designs an intervention to induce the target person's behavior to change to a target behavior (a behavior favorable to the user of the portable wireless communication device 100) via the already generated evaluation model (step S160). Thereafter, this intervention will be implemented as appropriate (Step S170).

[0032] The operation of the portable wireless communication device 100 will be explained below. The first program 131A, stored in the application section 131 of the external memory 130, functions as a thought reproduction system that reproduces the thoughts of a target person and is equipped with a thought reproduction AI (Sub-Belief System AI = SBS-AI) as a learning tool. The second program 131B functions as a behavior prediction system and is equipped with a behavior prediction AI (Prediction-AI) as a learning tool. Furthermore, the third program 131C functions as a behavior modification system and is equipped with a behavior modification AI (Behavior Modification AI) as a learning tool. Figure 5 is a conceptual diagram showing the collaborative relationship between the first to third programs 131A, 131B, and 131C.

[0033] As shown in Figure 5, in general terms, the target subject's biometric information (the target subject's response to external stimuli) and other information about the target subject (the target subject's behavioral information, and information about brain stimuli obtained from the external environment through the five senses and language) are input to the first program 131A (thought reproduction AI) as input information 220 (see Figure 3). Based on this input information 220, the first program 131A (thought reproduction AI) constructs an inference system for each target subject. This inference system reproduces the target subject's thought patterns or behavioral patterns, and the reproduced thought patterns or behavioral patterns are output 230 (see Figure 3). The second program 131B (behavioral prediction AI) takes the thought and behavior patterns of the target subject reproduced by the first program 131A as input 220, and based on these thought and behavior patterns, it predicts what actions the target subject will take afterward, or what kind of thought patterns (brain patterns) and behavior patterns they will develop after a specific period of time has elapsed, and outputs this prediction as output 230, thus constituting a behavioral prediction system.

[0034] This behavioral prediction forecasts the potential behavioral patterns of the target individual when they are in a natural state (peacetime situations, not emergencies such as war or disaster). Each behavior prediction AI is built as a pair with a thought reproduction AI, and when the thought reproduction AIs are parallelized, the same number of behavior prediction AIs are used to build the parallelization. The first program 131A and the second program 131B constitute a cognitive model 160 for the target audience. The third program 131C (behavior change system) takes the target subject's behavioral prediction output by the second program 131B (behavior prediction AI) as input 220. If the target subject's future behavior predicted by the second program 131B is undesirable, the third program 131C formulates an intervention plan (action plan) to change the target subject's behavior to a desirable behavior (target behavior), and outputs this intervention plan as output 230. The formulated intervention plan is implemented on the target subject in the real world to induce behavioral change in the target subject.

[0035] To build the behavioral change AI, we will use experimental data obtained from multiple subjects regarding intervention methods used in cognitive warfare. These three programs, 131A, 131B, and 131C, work together to form a cognitive-driven model that is fundamentally different from conventional event-driven models. Examples of intervention strategies formulated by the Third Program 131C (Behavioral Change System) include a variety of methods such as mass media, social media, verbal guidance, and liminal and subliminal effects through images and videos. In recent years, medical devices that directly affect the human body have also advanced, and methods have been developed to directly intervene in an individual's brain using advanced brain imaging devices such as functional magnetic resonance imaging (fMRI), electroencephalography (EEG), and magnetoencephalography (MEG).

[0036] As described above, in the portable wireless communication device 100, the thought-recreation AI analyzes the natural language messages of the target person and recognizes the target person's hidden intentions. For example, brain activation states measured in real time by fMRI (functional magnetic resonance imaging) are mutually fed back with the thought-recreation AI, and hidden intentions can be read from the patterns and the responses when intervening signals are sent to them. Behavioral prediction AI can generate hypotheses for multiple intentions, not just a single one. By pre-recognizing the brain activation state associated with each intention using fMRI or similar methods, it becomes possible to predict the thoughts and actions of target individuals. The following provides a detailed explanation of each of the three programs: the first program 131A (thought reproduction AI), the second program 131B (behavior prediction AI), and the third program 131C (behavior change system).

[0037] (A) First Program 131A (Thought Reproduction System, Thought Reproduction AI) Figure 6 is a conceptual diagram of the construction and updating of cognitive models by thought-reproducing AI (SBS-AI). The first program, 131A, constructs a "thought reproduction AI that represents brain patterns" as a single reasoning system for each target individual. The input 220 to the thought-reproducing AI (see Figure 3) includes, as shown in Figure 6, the biometric information of the target subject with a timestamp sent from the biometric information storage unit 132A, as well as the behavioral information of the target subject sent from the sub-data storage unit 132B, and stimulus information that the target subject receives from the external environment through the five senses and language. Representative examples of input 220, detection devices, and detection data are shown in Figures 7 to 9. Figure 7 shows an example of the target subject's biological information, Figure 8 shows an example of the target subject's behavioral information, and Figure 9 shows an example of the stimulus information the target subject receives from the external environment.

[0038] The cognitive model 160, composed of the first program 131A (thought reproduction AI), has the function of numerically analyzing correlations from input information 220 and identifying the thought patterns and behavioral patterns of the target individual. Numerical analysis methods include principal component analysis, which is used in traditional statistics; support vector machines, which are used in machine learning; and deep learning and large-scale language models (LLMs), which are statistical processing (AI) methods using large datasets. Because analytical methods are constantly evolving, it is preferable to adopt the most appropriate method available at the time. In the first program, 131A (thought reproduction AI), for example, a neural network is used. The thought-recreation AI is updated by applying the backpropagation method using the continuously received input 220. By continuously implementing this feedback, the cognitive model 160 of the target subject can be kept up-to-date. There are several ways to continuously acquire input information (220).

[0039] For example, it is possible to acquire data on target individuals through portable wireless communication devices (smartphones, wearable devices, small sensors) used by the target individuals themselves. Alternatively, data on the target individual can be obtained from IoT devices or smart home devices installed near the target individual (in their home or office). Alternatively, data can be obtained by tracking the online activities of target audiences, such as their social media usage, online search behavior, and website browsing history. Alternatively, you can collect opinions and feedback directly from your target audience through online surveys or in-app pop-up surveys.

[0040] Furthermore, by using devices equipped with biosensing technology (for example, Apple Vision Pro®), it is possible to directly acquire biometric information from target individuals, such as brain waves, muscle activity, and eye movements. In the real world, all information surrounding a target individual (symbolic information such as language, quasi-symbolic information such as subliminal messages, and non-symbolic information such as biofeedback) triggers neural information feedback at all levels of abstraction, thus updating the target individual's cognition. This is a phenomenon called homeostatic synchronization, which arises from the interaction between internal representations and the external environment, and is something that living organisms possess innately. Therefore, the method of updating a thought-reproducing AI using backpropagation as result data from input information can be said to be a logic that mimics the homeostatic synchronization phenomenon in a computer. As a result, by continuously updating the thought-reproducing AI, it is possible to bring the cognitive characteristics of the target individual and the behavioral characteristics of the cognitive model 160 closer together.

[0041] To create an ideal cognitive model 160, it is preferable to create a thought reproduction system at the time of the target individual's birth and then input all subsequent biological, behavioral, and stimulus information. In this portable wireless communication device 100, 10 to approximately 10,000 thought-reproducing AIs are incorporated into a single cognitive model 160, arranged in parallel. Each of these thought-reproducing AIs represents a part of the target person's diverse brain patterns and personality. For example, brain patterns that are visually dominant, auditorily dominant, forgetful, high IQ, and strong genetic traits are each represented by a thought-reproducing AI. The assumption is that the target person always possesses multiple brain patterns, and that one of these brain patterns is activated depending on the environment and situation.

[0042] The parallelized thought-reproducing AIs, ranging from 10 to approximately 10,000, are all independent of each other. In each thought-reproducing AI, the weight coefficients of each neural network are calculated based on the input information (220), just as in the case of a single thought-reproducing AI. Similar to the case of a single thought-reproducing AI, each thought-reproducing AI can be updated using the backpropagation method based on continuous input 220. The activity level of each thought-recreation AI is evaluated based on the distance between the information (vector) about the target subject and the output vector from the thought-recreation AI. The closer the distance between the vectors, the easier it is to activate the AI, and the thought-recreation AIs are ranked in order of highest activity level. This distance is calculated using Euclidean distance or cosine similarity. At regular intervals, the activity level is re-evaluated using new input information, and the order of the thought-recreation AIs is updated.

[0043] (B) Second Program 131B (Behavior Prediction System, Behavior Prediction AI) Figure 10 is a conceptual diagram of the behavior prediction system composed of the second program 131B. The behavior prediction system is equipped with a behavior prediction AI, which is connected to a thought reproduction AI and forms a pair with it. The behavior prediction AI predicts the thoughts and actions of a target person over a time interval T from a given time. The behavior prediction system takes the thought patterns or behavioral patterns of the target subject output from the first program 131A (thought reproduction AI) as input 220 (see Figure 3), numerically analyzes the temporal correlation from input 220, and has the function of predicting the future thoughts and behaviors of the target subject. Numerical analysis methods include methods that handle the temporal correlation function using statistics and machine learning, as well as statistical processing (AI) using large datasets such as DL and Large-Scale Language Models (LLM). Given the remarkable advancements in analytical methods, it is preferable to adopt the most appropriate method available at the time. For example, this behavior prediction system employs a method using neural networks.

[0044] The time interval T is not limited to one; multiple intervals can be selected. When selecting multiple intervals T, the order of selection is arbitrary, but typical examples include five intervals: 10 seconds later, 1 hour later, 1 week later, 1 month later, and 3 months later. Generally, shorter time intervals T are more suitable for predicting specific individual behavior, while longer time intervals T are more suitable for predicting organizational trends. By applying backpropagation to the behavior prediction AI based on continuous input information, the weight coefficients of the neural network are updated at each time interval T of the future prediction, thereby updating the behavior prediction system. The behavior prediction system calculates between 10 and approximately 100 predicted behaviors that will occur after a time interval T.

[0045] If there are 10 predicted actions, then for one target individual, using a pair of thought-reproducing AI and action-predicting AI, 10 actions will be predicted for each of the 5 time intervals (10 seconds later, 1 hour later, 1 week later, 1 month later, and 3 months later), resulting in a total of 1 × 5 × 10 = 50 predicted actions. The most appropriate predictive action is selected from these 50 predictive actions. When selecting the most appropriate predictive action, two indicators may be used, for example, "Most Likely CoA (MLCoA)" and "Most Dangerous CoA (MDCoA)." One method for calculating the "Most Probable Action (MLCoA)" is to use a softmax function in the output layer to calculate the probability of each action. One method for calculating the "Most Risky Behavior (MDCoA)" is to first define criteria for quantitatively evaluating the risk of each behavior, create a database of statistical risk assessments based on past data, and then use this database to calculate the risk of the predicted behavior. A corresponding behavioral prediction AI is built for each thought-reproducing AI, and the two are connected. This generates a unique prediction pattern for each thought-reproducing AI.

[0046] Similar to a single behavior prediction system, each behavior prediction AI is updated using backpropagation based on continuous input information. As mentioned above, each thought-reproducing AI creates 10 to 100 predicted actions for each time interval T, resulting in a total of 10 × 5 × 10 = 500 predicted actions. From these 500 predictive behaviors, the most appropriate predictive behavior is selected based on the two criteria mentioned above: "Most Probable Behavior (MLCoA)" and "Most Risky Behavior (MDCoA)". From the perspective of analytical philosophy, it is said that "for every 10 minutes in a person's life, there are 10 possible worlds."

[0047] A cognitive model composed solely of thought-reproducing AI can only make inferences based on the target individual's past and present information. In contrast, a cognitive model composed of a pair of thought-reproducing AI and behavior-predicting AI incorporates the element of time progression, thus strengthening the inference results. Furthermore, if we interpret "brain patterns" as the activated brain regions and "thought patterns" as emotions (or feelings), then the following correlation exists between the two. TIFF0007915502000001.tif57141

[0048] Furthermore, when using images to trigger these thought patterns (emotions), the following are examples of images. TIFF0007915502000002.tif78152

[0049] (C) Third Program 131C (Behavior Change System, Behavior Change AI) Figure 11 is a conceptual diagram of the behavior change system (behavior change AI) composed of the third program 131C. The behavior change system is an inference system that, when a target behavior (target behavior) is set for a specific target individual based on the thought reproduction system and behavior prediction system described above, designs intervention measures to induce the target individual to change their behavior in that manner, such as "we want this target individual to perform a specific action (target behavior) after a specific period of time." One behavioral change system is built for each target individual. In other words, one behavioral change AI is built for each cognitive model of the target individual, and it is not parallelized like thought reproduction AI or behavioral prediction AI. The input information for building the behavioral change AI will be experimental data from various intervention behaviors used in cognitive warfare, which were administered to subjects. This experimental data includes information about the stimuli that affected the subjects, as well as their behavior and biological information after receiving the stimuli.

[0050] Intervention methods include mass media and social media, verbal guidance, and liminal and subliminal effects through images and videos. It is also possible to use medical devices that directly affect the human body. Brain imaging devices such as functional magnetic resonance imaging (fMRI), electroencephalography (EEG), and magnetic encephalography (MEG) can be used to directly intervene in the target person's brain. When initially building a behavioral change AI, limiting the target to specific individuals would result in limited information. Therefore, it is possible to build a behavioral change AI based on a broad range of subject data and then customize this generalized behavioral change AI to one specifically for the target group. By copying a generalized behavioral change AI, it is possible to build a behavioral change AI that is specialized for each target individual. After the behavioral change AI is put into operation, the weight coefficients of the behavioral change AI, which are specific to the target group, will be continuously updated using new input information from when intervention behaviors are implemented on the target group.

[0051] As mentioned above, the behavioral change AI works in collaboration with the thought reproduction AI and the behavioral prediction AI to design intervention strategies to guide target individuals toward target behaviors. The behavioral change AI takes the "target individual, the desired behavior (target behavior), and its deadline" output from the second program 131B (behavioral prediction AI) as input 220 (see Figure 3) and designs intervention measures to encourage the target individual to perform the target behavior as output 230. The intervention strategies designed by the behavioral change AI do not need to be limited to just one; it is possible to design multiple intervention strategies depending on the situation of the target individual. Therefore, the design and simulation of 10 to 100 intervention strategies are carried out simultaneously.

[0052] Output 230 of the behavioral change AI shows the intervention plan, the specific actions to be taken as part of the intervention, and the timing of the intervention's implementation. Intervention strategies designed by behavioral change AI are placed in a simulation environment composed of thought-representation AI and behavior-prediction AI as input information, and are subjected to trial and verification. Through repeated trial and verification, the intervention strategies for the target individuals are optimized. Following this, intervention behaviors are implemented on the target individual, promoting a behavioral change from their natural behavior to the target behavior. When the intervention behaviors are implemented, neural information feedback occurs in the target individual at all levels of abstraction, including symbolic, quasi-symbolic, and non-symbolic information, and the target individual's cognition is altered (homeostatic synchronization). In this way, by continuously implementing intervention behaviors, the individual's cognitive functions in the real world gradually approach the behavior (target behavior) designed by the cognitive model 160 of the behavior change induction system 300.

[0053] (Intervention Delivery System) However, there are many challenges in implementing intervention measures. In particular, existing methods have limitations when implementing intervention measures for target groups. For example, in interventions targeting a wide range of people (target groups) through mass media, since each target individual has different brain patterns and other biological information, applying the same method to all target individuals will not yield uniform intervention effects. Furthermore, while the use of brain imaging devices is effective in interventions targeting a single individual, it lacks practicality because it requires the target individual's consent to use the device and also necessitates enormous computing resources to analyze the large amount of data.

[0054] Therefore, in order to solve these problems, the behavioral change induction system 300 according to this embodiment employs an intervention delivery system that enables it to respond to a large number of target individuals. Figure 12 is an overview diagram of the intervention delivery system. In this intervention delivery system, interventions are separated into abstract intervention policy signals and specific intervention actions. By separating intervention measures into intervention policy signals and intervention behaviors, it becomes possible to separate the AI ​​for behavioral change as well. For example, the behavioral change AI located on a cloud server could be responsible for designing intervention policy signals, while the intervention implementation AI located in an application on the target person's device (such as a mobile phone) could be responsible for generating and executing the intervention behavior.

[0055] Specifically, as shown in Figure 12, the intervention delivery system consists of a behavioral change AI 171 on a cloud server 170 that generates intervention policy signals, and an intervention implementation AI 174 on an application within a portable wireless receiver (e.g., a smartphone) 173 held by the target person 172. The behavioral change AI 171 (third program 131C) on the cloud server 170 collaborates with the first program 131A (thought reproduction AI) and the second program 131B (behavior prediction AI) to design intervention strategies for behavioral change in target individuals 172. The intervention strategies are encrypted and compressed into a lightweight intervention strategy signal of about 8 bytes, similar to MIDI data. This intervention strategy signal contains information with a certain degree of abstraction, and is configured to require a small amount of data. The intervention policy signal is transmitted via the existing information and communication infrastructure (Internet) to a portable wireless communication device 173 held by the target person 172.

[0056] When the intervention implementation AI 174 in the portable wireless communication device 173 receives an intervention policy signal from the behavior change AI 171, it formulates a customized intervention action for each target individual based on this intervention policy signal and puts it into action. Examples of these interventions include rewriting the text of chat messages, issuing fake earthquake alerts or other disaster warnings, and displaying subliminal images on the displays of portable wireless communication devices 173. Furthermore, it is also possible to obtain information about the target person 172 (such as operation history, browsing history, and center information) through the portable wireless communication device 173 via the application (intervention AI 174). For example, if the above intervention delivery system is used in the marketing field, the intervention can be divided into an intervention policy signal, such as "get potential customers to view the company's product campaign page within one hour," and an intervention action, such as "displaying a push notification on the display of the potential customer's portable wireless communication device (smartphone, etc.)" in order to achieve that objective.

[0057] Furthermore, when the intervention strategy delivery system is used in the military domain, the intervention strategy can be divided into two parts: an intervention policy signal that aims to "persuade the enemy leader to abandon a military invasion within six months," and an intervention action that involves "making an official announcement in international organizations to create international public opinion that a military invasion cannot be justified" in order to achieve that objective. Figures 13, 14, and 15 show examples of inputs and outputs (intervention policy signals and intervention actions) in the security, marketing, and entertainment domains, respectively. [Examples]

[0058] The following are examples of more specific applications of the behavioral change induction system 300 according to this embodiment.

[0059] (First example) In the first embodiment, the behavioral change induction system 300 is configured as a cognitive warfare defense system. In cognitive warfare defense systems, targets include, for example, the leaders of enemy countries, their military personnel, their citizens, and the citizens of other countries (one or more) who are key players in international public opinion. If computing resources are sufficient, it is preferable to build a cognitive model for the entire population; however, if computing resources are limited, the target group should be limited to one or a few people. For example, the leaders of enemy nations would be limited to the prime minister and several close advisors who have influence over the prime minister's decisions, and the military personnel of enemy nations would be limited to the commander of the enemy army and close advisors who have influence over the commander's decisions. Furthermore, for citizens of enemy nations and nationals of other countries who are the main actors in international public opinion, cognitive models would be constructed based on collective perceptions such as those belonging to a certain age group, residents of a specific region, or citizens of a specific country.

[0060] For example, if the target action against an enemy country is to prevent war, examples of intervention policy signals include making the enemy country's leaders feel that they will not be able to maintain their positions if they start a war, making the enemy country's military personnel feel that they have no chance of winning from a military and logistical standpoint, and making the enemy country's citizens feel that their lives will become difficult if war breaks out. Furthermore, regarding the citizens of other countries who are the main actors in international public opinion, this could include encouraging them to raise their voices in condemnation of the enemy country's declaration of war.

[0061] (Second example) In the second embodiment, the behavior change induction system 300 is configured as a marketing system. In marketing systems, the goal is to understand consumer behavior and expand influence based on that understanding. The behavior change induction system 300 reproduces consumers' decision-making processes, preferences, needs, and behavioral patterns, making it possible to design and implement personalized marketing strategies tailored to target audiences (potential customers). This marketing strategy involves, for example, the following steps: Firstly, a cognitive model 160 (see Figure 5) for each consumer (target group) is constructed. Specifically, individual cognitive models for each consumer are constructed based on biometric information such as consumers' brain patterns, as well as information related to purchasing behavior, such as each consumer's past purchase history, online behavior trajectory, and reactions and opinions expressed on social media.

[0062] This cognitive model 160 allows us to infer what cognitive biases consumers have towards specific products and services, or what factors stimulate or inhibit their purchasing intent. Secondly, personalized marketing strategies are designed. Specifically, using the constructed cognitive model 160, marketing messages and promotional activities optimized for consumers' current needs and future desires, as suggested by the behavioral prediction AI (second program 131B), are planned. This allows for the reproduction of consumers' deep psychology and latent needs in the marketing domain, creating effective approaches that directly lead to improved brand value and increased sales for the company.

[0063] (Third example) In the third embodiment, the behavioral change induction system 300 is configured as an entertainment system. By applying the behavioral change induction system 300 to the entertainment domain, it is possible to provide personalized experiences tailored to the individual preferences and behavioral patterns of a wide range of target audiences, including moviegoers and game and music enthusiasts. In this entertainment system, for example, the following steps are performed: Firstly, cognitive models 160 for each viewer and other target audience are constructed. Specifically, in addition to biometric information such as brain patterns, heart rate, skin conductivity, electroencephalogram (EEG), and pupillary response when viewers watch a work, an individual cognitive model 160 is constructed for each target individual based on their past movie viewing history, game play records, and music streaming information. This cognitive model 160 makes it possible to infer which genres and themes each target individual is particularly interested in, what types of stories and gameplay they prefer, and what moves them emotionally.

[0064] Secondly, personalized content will be recommended. Using a constructed cognitive model, the system recommends the most suitable movies, games, music, and other content for each target audience. Furthermore, it can recommend content at a detailed level, including visual presentation, game story progression, and musical tone. Thirdly, the interactive content experience is customized based on the viewer's cognitive model. For example, the system analyzes movie scenes that viewers have previously reacted to emotionally, recommending content that evokes similar emotions, or adjusting in-game events to match the player's preferences. Fourthly, the user experience is continuously optimized. By incorporating feedback from target audiences and new behavioral data into cognitive models, the target audience's experience is continuously optimized. This allows for rapid response to users' changing preferences and emerging trends, ensuring that the best possible entertainment experience is always provided.

[0065] In this way, by applying the behavioral change induction system 300 to the entertainment domain, the entertainment industry can provide more personalized and immersive experiences, significantly improving viewer and player satisfaction and loyalty. Furthermore, this approach can also be expected to promote the discovery of new content and expand entertainment consumption. To facilitate understanding of the behavioral change induction system 300 according to this embodiment, specific examples are given below.

[0066] (Specific example 1) 1. Target audience Individual A will be the target person. 2. Users of this behavioral change induction system 300 A friend or family member of individual A shall use this behavior change induction system 300 to attempt to change individual A's behavior. 3. Training data The following input data (stimulus) and output data (response) will be used as training data 200 (see Figure 3). The input data (stimuli) given to individual A is as follows: (A) Show scary images (movies, television, etc.) (B) Tell scary stories (such as reading thriller novels aloud) The responses (output data) shown by individual A to these input data (stimuli) were as follows: (1) The heart rate increases (2) To break out in a cold sweat (3) Trying to escape

[0067] These training data 200, along with other biometric information of individual A, are sent to the evaluation model generation program 135B by the training data input program 135A of the first program 131A, and the evaluation model generation program 135B generates an evaluation model 210 using machine learning (see step S120 in Figure 4). 4. Behavioral prediction using behavioral prediction AI After the evaluation model 210 has undergone training, the second program 131B (behavior prediction AI) predicts what action individual A will take in response to real-world stimuli. When individual A encounters the following stimuli (a) through (c), the second program 131B (behavior prediction AI) makes the following behavioral prediction. (a) Person A listens to a ghost story

[0068] The behavioral prediction AI predicts that individual A's response to this stimulus will be either "no reaction" or "a slight increase in heart rate." (b) While walking down a street at night, person A mistook something for a ghost. The behavioral prediction AI predicts that individual A's response to this stimulus will be "fall into a state of mental paralysis (their mind goes blank)." (c) While walking in the mountains, a snake suddenly appeared in front of me. The behavioral prediction AI predicts that individual A's response to this stimulus will be either "unable to move their legs to escape" or "becoming paralyzed with fear." 5. Behavioral change actions by behavioral change AI Based on the behavioral predictions for individual A output by the behavioral prediction AI, the third program 131C (behavioral change system) formulates behavioral change actions to modify individual A's behavior.

[0069] There are two directions to behavioral change. (1) Behavioral changes that weaken the emotions (affects) caused by a stimulus (debilitating behavioral changes) (2) Behavioral changes that replace emotions (feelings) caused by a stimulus with other emotions (substitutional behavioral changes) The behavioral change actions formulated by the Third Program 131C (Behavioral Change System) are one or both of these. (5.1) For example, consider a case where individual A has to go out in the middle of the night. In this case, the detrimental behavioral change action formulated by the behavioral change AI (a behavioral change action for "person A mistook something for a ghost") might be something like this:

[0070] (A) Give individual A a helmet with a powerful light, have them wear the helmet, and walk while illuminating the path ahead. If the area ahead is well-lit, the likelihood of "individual A mistaking something for a ghost" decreases significantly. The third program 131C (Behavior Change System) notifies the user of the Behavior Change Induction System 300 (a friend or family member of individual A) of this behavioral change action, for example, via email, or displays it on a screen. This allows the user of the Behavior Change Induction System 300 to purchase a helmet with a powerful light and have individual A carry it. Alternatively, if time permits, the third program 131C (behavior change system) can automatically order a helmet with a light from a product sales website (for example, an online shop such as Amazon®) and have it delivered via express delivery to the user of the behavior change induction system 300, or directly to individual A.

[0071] (B) The conversation continues between the automated voice and individual A via individual A's portable wireless communication device. Specifically, the third program 131C (behavior modification system) analyzes the words spoken by individual A, generates corresponding words via the automated voice system, and has them spoken from individual A's portable wireless communication device. This forces individual A to concentrate on the conversation, leaving them no time to mistake anything for a ghost. Alternatively, an icon could appear on the display of individual A's mobile phone, and individual A could then engage in a conversation with that icon. (C) Third Program 131C (Behavior Modification System) uses the map function of Individual A's portable wireless communication device to find a well-lit route to the destination and guides Individual A to the destination using Individual A's portable wireless communication device. Avoiding dark roads can reduce the possibility of misidentification. Alternatively, instead of a well-lit route, it would be possible to adopt a route with more foot traffic. The presence of people creates a sense of security, which reduces the possibility of misidentification, a common cause of anxiety.

[0072] Next, the replacement behavioral change actions formulated by the behavioral change AI might include, for example, the following: (D) Play beautiful nighttime city scenery and soothing music to individual A's portable wireless communication device to evoke positive emotions. This not only alleviates individual A's fear but also brings out positive emotions in individual A through enjoying beautiful scenery and soothing music. (E) Encourage video calls with friends and family via Personal A's portable wireless communication device, thereby providing Personal A with a sense of security and fostering a sense of human connection and solidarity. In situations where individual A would normally feel fear, the system helps individual A feel secure by reminding them of their connections with friends and family.

[0073] (F) Display photos and videos of past happy memories on the display of Person A's portable wireless communication device to evoke positive emotions. In situations where individual A would normally feel fear, recalling pleasant past memories overwrites individual A's fear with positive emotions. (5.2) For example, let's consider a case where individual A decides to go hiking in the mountains. In this case, the detrimental behavioral change action formulated by the behavioral change AI (a behavioral change action in response to "a snake suddenly appearing in front of you") would be as follows: (A) Have individual A wear sturdy (highly impact-resistant) boots. Since most snake bites occur on the feet, this is intended to give individual A a sense of security, knowing that protecting their feet will ensure they are safe.

[0074] The third program 131C (behavior change system) notifies the user of the behavior change induction system 300 (a friend or family member of individual A) of this behavior change action, for example, via email, or displays it on a screen. This allows the user of the behavior change induction system 300 to purchase rain boots and have individual A wear them. Alternatively, if there is sufficient time, the third program 131C (behavior change system) can take measures such as automatically ordering rubber boots from a product sales website (for example, an online shop such as Amazon®) and having them delivered by express delivery to the user of the behavior change induction system 300 or to individual A.

[0075] (B) Have individual A wear clothing that prevents heat dissipation. Snakes use their tongues to sense heat and recognize their prey, so this is to prevent heat from being released to the outside. The third program 131C (behavior change system) notifies the user of the behavior change induction system 300 (a friend or family member of individual A) of this behavior change action, for example, via email, or displays it on a screen. This allows the user of the behavior change induction system 300 to purchase heat-insulating clothing and have individual A wear it. Alternatively, as in case (A), it is also possible to place an automated order on a website selling goods. (C) The third program 131C (behavior modification system) provides the suggestion to individual A via the display 142 or speaker 145 of individual A's portable wireless communication device, that is, visually or audibly, that "snakes are not scary." In most cases, when a human encounters a snake, the snake will run away unless the human attacks, so this is to prevent individual A from feeling excessive fear.

[0076] Next, the replacement behavioral change actions formulated by the behavioral change AI might include, for example, the following: (D) Display positive information about the beautiful scenery and flora and fauna of the mountains on the display of Person A's portable wireless communication device, so that Person A can enjoy them while walking. By overwriting the fear of snakes with information about beautiful scenery and interesting flora and fauna, the fear of snakes is alleviated, and people can enjoy hiking in the mountains. For example, a user of the behavioral change induction system 300 can send instructions to the third program 131C (behavioral change system) of individual A's portable wireless communication device via individual A's portable wireless communication device, and these instructions can be displayed.

[0077] (E) Audio guides emphasizing the health and relaxation benefits of hiking are played from individual A's portable wireless communication device. Overwrite your fear of snakes with the positive effects of health and relaxation. For example, a user of the behavioral change induction system 300 can send instructions to the third program 131C (behavioral change system) of the personal A's portable wireless communication device via the personal A's portable wireless communication device, and these instructions can be played. (F) Display photos and videos that evoke memories of past enjoyable hikes and nature explorations on the display of Person A's portable wireless communication device. In situations where individual A would normally feel fear, they overwrite their fear of snakes with positive emotions by replaying past enjoyable hiking or nature exploration experiences. For example, a user of the behavioral change induction system 300 can send instructions to the third program 131C (behavioral change system) of individual A's portable wireless communication device via individual A's portable wireless communication device, and these instructions can be displayed.

[0078] (Specific example 2) 1. Target audience The target is President P of Country R. 2. Users of this behavioral change induction system 300 The government of country F, which is in conflict with country R, ​​will use this behavioral change induction system 300 to attempt to change the behavior of President P of country R. 3. Training data The following input data (stimulus) and output data (response) will be used as training data 200 (see Figure 3).

[0079] The input data (stimuli) to be given to the target individuals is as follows: (A) The domestic economy is sluggish (the price of crude oil, which is a source of foreign currency for Country R, ​​is falling, and foreign currency is not flowing into Country R). (B) A large new oil field is discovered in another country, or a shale revolution occurs in another country (this will further lower crude oil prices, making the economy of country R, ​​which is dependent on crude oil, even more difficult). The responses (output data) shown by the target subjects to these input data (stimuli) were as follows: (1) To divert public discontent caused by a weak domestic economy, military activities are conducted abroad (such as annexing the territory of another country or militarily intervening in a conflict in another country). (2) Strengthen the domestic intelligence network (to expose disgruntled elements within the country) (3) Strengthen military power

[0080] 4. Behavioral prediction using behavioral prediction AI After the evaluation model 210 has undergone training, the second program 131B (behavior prediction AI) predicts how the target person will behave in response to real-world stimuli. When the target individual encounters the following stimuli, the second program 131B (behavior prediction AI) makes the following behavioral predictions. (a) Country S, which had maintained neutrality toward both Country R and its enemy Country F, abandons its previous neutrality policy and joins the military alliance N (e.g., NATO) to which Country F belongs. The behavioral prediction AI predicts that the target individual will respond to this stimulus by "invading neighboring country U, which is not yet a member of military alliance N, and annexing neighboring country U to country R, ​​in order to prevent further expansion of military alliance N."

[0081] 5. Behavioral change actions by behavioral change AI Based on the behavior prediction of the target individual (invasion of neighboring country U) output by the behavior prediction AI, the third program 131C (behavior change system) formulates behavior change actions to alter the target individual's behavior. As mentioned above, the third program 131C (Behavior Change System) develops both detrimental behavior change actions and substitutional behavior change actions. The detrimental behavioral change actions formulated by the behavioral change AI might include, for example, the following: (A) We will request that the leaders of each country strengthen their military support for the United Nations. Specifically, the third program 131C (Behavior Change System) encourages Country F, a user of this Behavior Change Induction System 300, to issue a statement or send an email to the heads of state of other countries. (B) Encourage the media to raise international public awareness. For example, the third program 131C (Behavior Change System) involves country F, a user of this behavior change induction system 300, providing information about country R's actions and intentions to media outlets in various countries and encouraging them to publish articles protesting and condemning country R's planned invasion of country U.

[0082] (C) We foreshadow economic sanctions (boycott of oil from Country R). For example, the third program 131C (Behavioral Change System) encourages Country F, a user of this Behavioral Change Induction System 300, to issue a statement threatening economic sanctions, or to issue a notice to the heads of state of other countries urging them to participate in economic sanctions. (D) The countries hostile to Country R will emphasize their support for Country U, and these countries will issue statements to foster a sense of international solidarity. This would make the target feel internationally isolated, fear diplomatic failure, and thus deter them from invading the U.S. For example, the third program 131C (Behavior Change System) encourages Country F, a user of this Behavior Change Induction System 300, to notify the heads of state of other countries of the holding of an international conference for issuing a statement, or to encourage Country F to request that various national media outlets publish articles appealing to the need for an international conference.

[0083] (E) Propose an international economic cooperation program to support the economic recovery of Country R. As Country R's economy recovers, the need to invade Country U to divert its citizens' attention abroad will gradually diminish. For example, the third program 131C (Behavior Change System) encourages Country F, a user of the Behavior Change Induction System 300, to present such a policy and issue a notice calling for cooperation from other countries. Furthermore, it encourages Country F to request that national media outlets publish articles highlighting the need for such a program. (F)R will conduct an information war to suppress anti-government activities within the country and emphasize domestic political stability. If the domestic political situation stabilizes, the need to invade Country U to divert the public's attention abroad will decrease. For example, the third program 131C (Behavior Change System) presents such a policy to Country F, a user of the Behavior Change Induction System 300, and encourages it to issue a notice calling for cooperation from other countries (allies).

[0084] Next, the replacement behavioral change actions formulated by the behavioral change AI might include, for example, the following: (G) Provide international conferences and negotiation venues that highlight the diplomatic success of the target group and foster a sense of well-being from diplomatic success. By emphasizing international cooperation and support for Country R, ​​and making the target feel a sense of diplomatic success, the motivation for invasion will be reduced. For example, the third program 131C (Behavior Change System) encourages Country F, a user of this Behavior Change Induction System 300, to notify the heads of state of other countries about the holding of the international conference, or to ask the media in various countries to publish articles emphasizing the necessity of holding the international conference. (H)R Propose economic policies and infrastructure projects that will lead to an increase in the approval rating of the target person (President P) within the country, and elicit a sense of happiness from the target person as their approval rating increases. By emphasizing domestic economic stability and infrastructure projects, and gaining public support, they can make the need for invasion seem less compelling. For example, the third program 131C (Behavior Change System) presents such policies and projects to Country F, a user of the Behavior Change Induction System 300, and encourages Country F to communicate them to Country R.

[0085] (I) Promote the hosting of cultural and sporting events that will earn the target audience recognition both domestically and internationally, thereby building a positive international image of the target audience. Through international sporting events and cultural exchanges, the target audience will be made to feel a positive international image and admiration, thus eliminating the need for invasion. For example, the third program 131C (Behavior Change System) proposes to Country F, a user of the Behavior Change Induction System 300, the plan for such a sporting event, and encourages it to notify heads of state of affairs of their participation in the sporting event, or to request that media outlets in various countries publish articles emphasizing the need to hold the sporting event.

[0086] (J) Provide new diplomatic initiatives that will make target audiences feel internationally successful. By making the target feel internationally supported and emphasizing diplomatic successes, the goal is to alleviate their fear of downfall and discourage them from invading the United States. For example, the third program 131C (Behavior Change System) encourages Country F, a user of the Behavior Change Induction System 300, to notify heads of state of other countries of the holding of an international conference to offer diplomatic initiatives to Country R, ​​or to request that national media outlets publish articles emphasizing the need for such an international conference.

[0087] (K) Provide success stories and positive data that emphasize the economic growth and domestic stability of Country R, ​​creating a sense of economic stability. The goal is to alleviate economic fears and overwrite them with data that conveys domestic stability and growth, thereby discouraging an invasion of the U.S. For example, the third program 131C (behavior change system) presents such cases and data to country F, which is a user of the behavior change induction system 300, and has country F transmit them to the target individuals. (L) Provide examples of successful cultural exchanges and sporting events that have garnered international acclaim for the target audience, thereby emphasizing a positive international image. Through international success stories, the aim is to elicit positive emotions from target individuals and alleviate their fears, thereby deterring an invasion of the U.S. For example, the third program 131C (behavior change system) presents emails and videos demonstrating such cases to country F, a user of the behavior change induction system 300, and country F sends these emails and videos to media outlets in various countries.

[0088] (Specific example 3) 1. Target audience The target audience is heavy viewers of videos on portable wireless communication devices (smartphones) who rarely go to movie theaters. 2. Users of this behavioral change induction system 300 A friend or family member of the target individual shall use this behavioral change induction system 300 to attempt to change the target individual's behavior. 3. Training data The following input data (stimulus) and output data (response) will be used as training data 200 (see Figure 3). The input data (stimuli) to be given to the target individuals is as follows: (A) Show your favorite short video clip (B) Have them watch the video while playing relaxing music. (C) Providing video content that can be enjoyed casually. The responses (output data) shown by the target subjects to these input data (stimuli) were as follows: (1) Show a smile (2) Take a relaxed posture (3) Feeling comfortable

[0089] 4. Behavioral prediction using behavioral prediction AI After the evaluation model 210 has undergone training, the second program 131B (behavior prediction AI) predicts how the target person will behave in response to real-world stimuli. When the target individual encounters the following stimuli, the second program 131B (behavior prediction AI) makes the following behavioral predictions. (a) Show the target audience trailers for action movies that will excite them. The behavioral prediction AI predicts that the target individual will respond to this stimulus by either "showing an excited expression" or "leaning forward." (b) A friend makes plans to go to the movies with the target person. The behavioral prediction AI predicts that the target individual will respond to this stimulus by "actively stating, 'I want to go to the movie theater.'" (c) Show the target audience information about special events at movie theaters. The behavioral prediction AI predicts that the target individual will respond to this stimulus by either "deciding to go to the movie theater" or "starting to prepare to go to the movie theater."

[0090] 5. Behavioral change actions by behavioral change AI Based on the behavioral predictions of the target individuals output by the behavioral prediction AI, the third program 131C (behavioral change system) formulates behavioral change actions to alter the target individuals' behavior. As mentioned above, the third program 131C (Behavior Change System) develops both detrimental behavior change actions and substitutional behavior change actions. (5.1) For example, consider a case where the target person wants to relax at home and watch videos on their portable wireless communication device (smartphone). In this case, the detrimental behavioral change actions formulated by the behavioral change AI might be something like the following:

[0091] (A) Limit the amount of time that can be spent watching videos on portable wireless communication devices. For example, the third program 131C (behavior change system) sends a reminder to the user of the behavior change induction system 300 that the target person's video viewing time should be limited, and displays a message to that effect on the display screen of the target person's portable wireless communication device, or periodically plays an automated voice message via the target person's portable wireless communication device. Alternatively, the user of this behavioral change induction system 300 can remotely control the target person's portable wireless communication device to forcibly limit the target person's video viewing time. (B) Inform viewers about the health effects of prolonged viewing. For example, the third program 131C (behavior change system) notifies the user of the behavior change induction system 300, and the user directly informs the target person. Alternatively, it is possible to display a message on the screen of the target person's portable wireless communication device, or to periodically play an automated voice message.

[0092] (C) After watching a video for a certain period of time, a message is sent prompting the user to take a break. For example, the third program 131C (behavior change system) notifies the user of the behavior change induction system 300, prompting the user to encourage the target person to take a break. Alternatively, it is possible to display a message on the screen of the target person's portable wireless communication device, or to periodically play an automated voice message. Next, the replacement behavioral change actions formulated by the behavioral change AI might include, for example, the following: (D) By showing trailers and making-of videos of exciting action movies, the target audience will be made to expect that seeing the movie in a theater will be even more thrilling. For example, the third program, 131C (Behavior Modification System), displays such images on the target person's portable wireless communication device, replacing watching videos at home with the stimulating experience of going to a movie theater.

[0093] (E) Notify people about special events and limited screenings at movie theaters and encourage them to plan trips to the movies with friends. For example, the third program 131C (behavior change system) notifies the user of the behavior change induction system 300 of this fact, and the user then provides information to the target person. Alternatively, the third program 131C (Behavior Change System) could send an email to the target individual or inform them of this information via a display or speaker, emphasizing a special movie theater experience and encouraging positive behavior by prompting them to plan with friends. (F) Show the target audience photos and videos that evoke pleasant memories of going to the movies, and encourage them to want to go to the movies again. For example, the third program, 131C (Behavior Change System), displays such photos and videos on the display of the target person's portable wireless communication device, thereby reminding them of past enjoyable experiences at movie theaters and stimulating their behavior.

[0094] (5.2) For example, consider a scenario where the target audience decides to go to a movie theater. In this case, the detrimental behavioral change actions formulated by the behavioral change AI might be something like the following: (A) Provide an application that allows users to easily search for and book transportation to the movie theater. For example, the third program 131C (Behavior Change System) automatically downloads such an application to the target person's portable wireless communication device without any instruction from the target person or user, thereby reducing the hassle of checking travel routes and lowering the barrier to the target person going to the movie theater. (B) Offer discount coupons and special offers for movie theaters to highlight the benefits of going to the movies. For example, the third program, 131C (Behavior Change System), displays discount coupons and special offers for movie theaters on the display of the target person's portable wireless communication device. This creates a feeling that there is something to be gained by going to the movie theater, thereby encouraging the target person to take action.

[0095] (C) Provides real-time information on movie theater crowd levels and showtimes to help you go at the best time. For example, the third program 131C (Behavior Change System) displays this information on the target person's portable wireless communication device, making them expect a smoother movie theater experience and encouraging their behavior. Next, the replacement behavioral change actions formulated by the behavioral change AI might include, for example, the following: (D) Emphasize special events and interactive experiences at movie theaters, suggesting that there are unique pleasures that can only be found in movie theaters. For example, the third program, 131C (Behavioral Change System), displays this information on the screen of a portable wireless communication device, replacing watching videos at home with the special experience of going to a movie theater.

[0096] (E) Suggest planning to enjoy a movie outing with friends and family, emphasizing social connections. For example, the third program 131C (Behavior Change System) displays such suggestions on the display of the target person's portable wireless communication device. It encourages the target person to engage in more active behavior by viewing the movie theater experience as a social event. (F) Provide video messages about past enjoyable experiences at movie theaters and new experiences that people can look forward to at movie theaters, thereby encouraging them to actively go to the movies. For example, the third program 131C (Behavior Change System) displays such video messages on the target person's portable wireless communication device, emphasizing positive experiences at the movie theater and encouraging the target person's behavior to become more active.

[0097] (Specific example 4) 1. Target audience The target is President P of Country R. 2. Users of this behavioral change induction system 300 The Japanese government will use this behavioral change induction system 300 to attempt to change the behavior of President P of Country R. 3. Training data The following input data (stimulus) and output data (response) will be used as training data 200 (see Figure 3). The input data (stimuli) to be given to the target individuals is as follows: (A) Show documents and records indicating that neighboring country U of country R has traditionally been part of country R's territory. (B) Information is received that residents of a specific region in neighboring country U are originally from country R, ​​and that these residents are being mistreated by the U government.

[0098] The responses (output data) shown by the target subjects to these input data (stimuli) were as follows: (1) Determined to help the residents of a specific region in neighboring country U, they decided to invade neighboring country U. 4. Behavioral prediction using behavioral prediction AI After the evaluation model 210 has undergone training, the second program 131B (behavior prediction AI) predicts how the target person will behave in response to real-world stimuli. When the target individual encounters the following stimuli, the second program 131B (behavior prediction AI) makes the following behavioral predictions. (a) I have come across documents and records indicating that the Ainu people, who currently reside in Hokkaido, Japan, were originally inhabitants of Country R and were the indigenous people of Japan. (b) The Ainu people are being mistreated by the Japanese government. The behavioral prediction AI predicts that the target individual will act in response to these stimuli (inputs) as follows: "By invading Hokkaido, Hokkaido will be annexed to Country R, ​​and the Ainu people will be saved."

[0099] 5. Behavioral change actions by behavioral change AI Based on the behavioral prediction of the target individual (invasion of Hokkaido) output by the behavioral prediction AI, the third program 131C (behavioral change system) formulates behavioral change actions to alter the target individual's behavior. As mentioned above, the third program 131C (Behavior Change System) develops both detrimental behavior change actions and substitutional behavior change actions. The detrimental behavioral change actions formulated by the behavioral change AI might include, for example, the following: (A) Send visible signs (issue a statement to the international community, raise an agenda item with the United Nations, strengthen the deployment of the Self-Defense Forces to Hokkaido). For example, the third program 131C (Behavior Change System) makes these proposals to the Japanese government, which is a user of this behavior change induction system 300. (B) Encourage the media to raise international public opinion. For example, the third program 131C (Behavior Change System) encourages the Japanese government, a user of this behavior change induction system 300, to provide the media with necessary information and request that they publish articles.

[0100] (C) The government announces economic sanctions (a boycott of oil from Country R) through statements and media. For example, the third program 131C (Behavior Change System) makes these proposals to the Japanese government, which is a user of this behavior change induction system 300. Next, the replacement behavioral change actions formulated by the behavioral change AI might include, for example, the following: (D) By supporting Country U, which is currently at war with Country R, ​​we will strengthen Country U's offensive and force Country R to abandon its invasion of Hokkaido. (E) Propose beneficial economic assistance to Country R, ​​thereby creating a reduction in tensions between your country and Country R. In either case, for example, the Third Program 131C (Behavior Change System) would formulate such a plan and propose it to the Japanese government, which is a user of the Behavior Change Induction System 300. [Explanation of Symbols]

[0101] 100 Portable wireless communication devices 110 Communications Department 120 Control Unit 130 External memory (hard disk) 131 Application Department 132 Data Storage Unit 140 Input / output section 150 antennas

Claims

1. A behavioral change induction system that designs intervention measures to induce behavioral change in target individuals, A storage means for storing biological information including the target person's response to past external stimuli, A thought pattern reproduction means that reproduces the thought patterns or behavioral patterns of the target person using a predictive model that has been machine-learned using the aforementioned biometric information as training data, A response estimation means takes current or expected future external stimuli to the target subject as input, and outputs the expected response of the target subject to those external stimuli based on the thought patterns or behavioral patterns of the target subject reproduced by the thought pattern reproduction means, An intervention design means for designing an intervention to guide the target person's behavior toward a target behavior, in accordance with the expected response of the target person output by the response estimation means, A behavioral change induction system equipped with the following features.

2. The behavioral change induction system according to claim 1, characterized in that the biological information includes brain pattern information of the target subject.

3. The behavioral change induction system according to claim 1, characterized in that the intervention design means designs at least one of the following: an intervention to guide the target person toward weakening the emotion caused by the external stimulus, and an intervention to guide the target person toward replacing the emotion caused by the external stimulus with another emotion.

4. A portable wireless communication device incorporating a behavioral change induction system according to any one of claims 1 to 3.

5. A behavior change promotion system that designs and implements intervention measures to induce behavioral change in target individuals, The behavior change promotion system comprises a behavior change induction system and a portable wireless communication device possessed by the target individual. The behavior change induction system is A storage means for storing biological information including the target person's response to past external stimuli, A thought pattern reproduction means that reproduces the thought patterns or behavioral patterns of the target person using a predictive model that has been machine-learned using the aforementioned biometric information as training data, A response estimation means takes current or expected future external stimuli to the target subject as input, and outputs the expected response of the target subject to those external stimuli based on the thought patterns or behavioral patterns of the target subject reproduced by the thought pattern reproduction means, An intervention plan design means for designing an intervention plan to guide the target person's behavior toward a target behavior, in accordance with the expected response of the target person output by the response estimation means, Equipped with, The aforementioned portable wireless communication device incorporates an application for designing intervention measures based on the aforementioned intervention policy. The behavioral change induction system transmits the intervention plan to the target person's portable wireless communication device. A behavior change promotion system in which the application of the portable wireless communication device that receives the intervention policy formulates customized intervention actions for each of the target individuals based on the intervention policy.

6. The behavioral change promotion system according to claim 5, characterized in that the aforementioned biological information includes brain pattern information of the target subject.

7. The behavior change promotion system according to claim 5, characterized in that the intervention design means designs at least one of the following: an intervention to guide the target person toward weakening the emotion caused by the external stimulus, and an intervention to guide the target person toward replacing the emotion caused by the external stimulus with another emotion.

8. A program for causing a computer to execute a behavioral change induction method, which involves designing intervention measures to induce behavioral change in target individuals. A first process that stores biological information including the target person's response to past external stimuli, A second process is performed to reproduce the thought patterns or behavioral patterns of the target person using a predictive model trained with the aforementioned biometric information as training data, A third process takes current or expected future external stimuli to the target subject as input, and outputs the expected response of the target subject to those external stimuli based on the thought patterns or behavioral patterns of the target subject reproduced in the second process. A fourth process involves designing an intervention strategy to guide the target person's behavior toward a target behavior, based on the expected response of the target person output in the third process. A program that causes a computer to execute something.

9. The program according to claim 8, characterized in that the fourth process is to design at least one of the following: an intervention to induce the target person to weaken the emotion caused by the external stimulus, and an intervention to induce the target person to replace the emotion caused by the external stimulus with another emotion.

10. A program for causing a computer to execute a behavioral change induction method, which involves designing intervention measures to induce behavioral change in target individuals. A first process that stores biological information including the target person's response to past external stimuli, A second process is performed to reproduce the thought patterns or behavioral patterns of the target person using a predictive model trained with the aforementioned biometric information as training data, A third process takes current or expected future external stimuli to the target subject as input, and outputs the expected response of the target subject to those external stimuli based on the thought patterns or behavioral patterns of the target subject reproduced in the second process. A fourth process involves designing an intervention strategy for guiding the target person's behavior toward a target behavior, based on the expected response of the target person output in the third process. A fifth process in which an application built into a portable wireless communication device owned by the target person formulates a customized intervention action for each target person based on the intervention policy, A program that causes a computer to execute something.

11. The program according to claim 10, characterized in that the fourth process is to design at least one of the following: an intervention to guide the target person toward weakening the emotion caused by the external stimulus, and an intervention to guide the target person toward replacing the emotion caused by the external stimulus with another emotion.

Images