Information processing methods

Abstract

Enable AI agents to perform tasks appropriately. [Solution] The information processing method is an information processing method in an agent system using artificial intelligence. The information processing method includes: a reception step in which the agent system receives user input; a selection step in which the agent system selects one inference procedure suitable for the received input from a predetermined number of types of inference procedures; and a processing step in which the agent system performs processing related to the received input according to a number of inference steps included in one inference procedure.

Description

Technical Field

[0001] The present invention relates to the technical field of information processing methods.

Background Art

[0002] As a method of this kind, for example, a method has been proposed in which an agent system provides information requested by a passenger in the vehicle interior while communicating with the passenger by voice (see Patent Document 1).

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] As an example of the above agent system, an AI (Artificial Intelligence) agent using AI has been proposed. The AI agent may autonomously execute tasks by utilizing available tools. However, there is a technical problem that the AI agent may not be able to appropriately utilize the tools. Note that in the AI agent, a learning model constructed by machine learning may be used.

[0005] The present invention has been made in view of the above problems, and an object thereof is to provide an information processing method capable of appropriately causing an AI agent to perform a task. [[ID=4०]]

Means for Solving the Problems

[0006] [[ID=४५]] An information processing method according to one aspect of the present invention is an information processing method in an agent system using artificial intelligence, wherein the agent system includes a reception step of receiving user input, a selection step of the agent system selecting one inference procedure suitable for the received input from a predetermined plurality of types of inference procedures, and a processing step of the agent system performing processing related to the received input according to a plurality of inference steps included in the one inference procedure. [Brief explanation of the drawing]

[0007] [Figure 1] This is a block diagram showing the configuration of the agent system according to the embodiment. [Figure 2] This is a block diagram showing an example of a computer configuration. [Figure 3] This is a flowchart showing the operation of the agent system according to the embodiment. [Figure 4] This figure shows an example of a prompt entered into the agent. [Figure 5] This figure shows an example of a prompt for the inference instruction tool according to the embodiment. [Modes for carrying out the invention]

[0008] Embodiments relating to the information processing method will be described with reference to Figures 1 to 5. In Figure 1, the agent system 1 comprises a server 10 and a terminal device 20 mounted on a vehicle 2. The server 10 and the terminal device 20 are configured to communicate with each other via a network NW. The server 10 may be implemented by a single server or by multiple servers. In other words, server 10 may represent a group of servers. Server 10 may also be a cloud server.

[0009] At least one of the server 10 and the terminal device 20 may be implemented by a computer COM as shown in Figure 2. The computer COM comprises an arithmetic unit 110, a storage device 120, a communication device 130, an input device 140, and an output device 150. The arithmetic unit 110, storage device 120, communication device 130, input device 140, and output device 150 are connected via a data bus 160.

[0010] The arithmetic unit 110 may have a processor. The arithmetic unit 110 may have a single processor or multiple processors. In other words, the arithmetic unit 110 may have one or more processors. Furthermore, the processor may be a multi-core processor. If the arithmetic unit 110 has a single processor that is a multi-core processor, then logically, the arithmetic unit 110 can be said to have multiple processors.

[0011] The processor may be at least one of the following: CPU (Central Processing Unit), GPU (Graphics Processing Unit), FPGA (Field Programmable Gate Array), and TPU (Tensor Processing Unit).

[0012] The storage device 120 may be at least one of the following: RAM (Random Access Memory), ROM (Read Only Memory), hard disk drive, magneto-optical disk drive, SSD (Solid State Drive), and optical disk array. In other words, the storage device 120 may be implemented by a single device or by multiple devices.

[0013] The communication device 130 may be capable of communicating with devices outside of the computer COM. Furthermore, the communication device 130 may use either wired or wireless communication.

[0014] The input device 140 is a device capable of receiving information input to the computer COM from an external source. The input device 140 may include an operating device (e.g., keyboard, mouse, touch panel, etc.) that can be operated by the user of the computer COM. The input device 140 may include a microphone capable of collecting the user's voice. In this case, the user may provide voice input. The input device 140 may include a recording medium reader capable of reading information recorded on a recording medium that can be attached to and detached from the computer COM, such as a USB (Universal Serial Bus) memory. Furthermore, when information is input to the computer COM via the communication device 130 (in other words, when the computer COM acquires information via the communication device 130), the communication device 130 may function as an input device.

[0015] The output device 150 is a device capable of outputting information to the outside of the computer COM. The output device 150 may have a display device capable of outputting visual information such as characters and images as the above information. The output device 150 may also have a speaker capable of outputting auditory information such as sound as the above information. The output device 150 may also have a vibration motor capable of outputting tactile information such as vibration as the above information. The output device 150 may also have a printer. The output device 150 may be capable of outputting information to a recording medium that can be attached to and detached from the computer COM, such as a USB memory stick. When the computer COM outputs information via the communication device 130, the communication device 130 may function as an output device.

[0016] The storage device 120 is capable of storing desired data. The storage device 120 may store the computer program CP that the arithmetic unit 110 will execute. The storage device 120 may temporarily store data that the arithmetic unit 110 will use temporarily when the arithmetic unit 110 is executing the computer program CP.

[0017] The arithmetic unit 110 (for example, a processor) may execute the processing that the computer COM should perform together with the memory device 120 in which the computer program CP is stored (in other words, together with the memory device 120 and the computer program CP stored in the memory device 120). For example, by the arithmetic unit 110 executing the computer program CP, a logical functional block for executing the processing that the computer COM should perform may be realized within the arithmetic unit 110 (for example, within the processor).

[0018] Returning to Figure 1, Server 10 has, as logical functional blocks, an AI agent 11, an inference instruction tool 12, a computation tool 13, and a Point of Interest (PoI) search tool 14. The inference instruction tool 12, the computation tool 13, and the PoI search tool 14 are external tools that can be used by the AI ​​agent 11. Server 10 may have external tools other than the inference instruction tool 12, the computation tool 13, and the PoI search tool 14. At least one of the inference instruction tool 12, the computation tool 13, and the PoI search tool 14 may be implemented by a server other than Server 10. In other words, the server that implements the AI ​​agent (in this case, Server 10) and the server that implements the external tools may be different. The AI ​​agent 11 may be an AI agent using generative AI such as a Large-Scale Language Model (LLM) or a Large-Scale Multimodal Model (LMM).

[0019] The operation of Agent System 1 will be explained with reference to the flowchart in Figure 3. User U of Vehicle 20 (in other words, the occupant of Vehicle 20) may provide input to AI Agent 11 via Terminal Device 20. Terminal Device 20 may transmit signals related to User U's input to Server 10 via Network NW. When Server 10 receives signals related to User U's input, the AI ​​Agent accepts User U's input (Step S101).

[0020] At this time, the AI agent may receive a prompt including the input content of the user U as shown in FIG. 4. Note that such a prompt may be generated by the terminal device 20. Note that in the prompt shown in FIG. 4, the prompting method called ReAct is used, but it is not limited thereto.

[0021] Next, the AI agent 11 calls the inference instruction tool 12 to select an inference procedure suitable for the input content of the user U (step S102). For example, as shown in FIG. 5, the prompt of the inference instruction tool 12 states, "This inference instruction tool must be called first." Therefore, the AI agent 11 calls the inference instruction tool 12 at the beginning of the process for generating an answer to the input of the user U.

[0022] As shown in FIG. 5, the prompt of the inference instruction tool 12 includes a plurality of "Patterns" corresponding to the above inference procedures. For example, Pattern1 is an inference procedure for proposing the one with the lowest price from a plurality of candidates. This Pattern1 includes four inference steps: "Search for candidates that meet the requirements.", "Calculate the price based on the conditions that meet the requirements.", "Sort by price criteria.", and "Answer with the cheapest facility along with the reason for recommendation."

[0023] Returning to FIG. 3, after the process of step S102, the AI agent 11 executes the process according to the selected inference procedure (step S103). Then, the AI agent 11 outputs the processing result (step S104). Specifically, the AI agent 11 may transmit a signal indicating the processing result to the terminal device 20 via the network NW. As a result, the processing result of the AI agent 11 (that is, the answer to the input of the user U) may be displayed on the screen of the terminal device 20.

[0024] For example, in the process of step S102, the AI ​​agent 11 may select Pattern 1 as the inference procedure. In this case, the AI ​​agent 11 may use the PoI search tool 14 to search for candidates in accordance with the inference step "Search for candidates that meet the requirements." Next, the AI ​​agent 11 may use the calculation tool 13 to calculate the price in accordance with the inference step "Calculate the price based on the conditions that meet the requirements." Next, the AI ​​agent 11 may sort the search results using the PoI search tool 14 by price in accordance with the inference step "Sort by price criterion." Next, the AI ​​agent 11 may output a response that includes the cheapest facility and the reason for the recommendation in accordance with the inference step "Recommend the cheapest facility along with the reason for the recommendation."

[0025] (Technical effects) An AI agent (for example, AI agent 11) may autonomously perform tasks by utilizing external tools. However, an AI agent may not be able to properly utilize external tools. Therefore, in this embodiment, an inference instruction tool 12 is introduced as an external tool. AI agent 11 selects an inference procedure from the inference instruction tool 12 that is appropriate for the input content of user U. Then, AI agent 11 performs processing (in other words, a task) according to the selected inference procedure. In this way, if AI agent 11 performs processing according to a predetermined inference procedure, AI agent 11 can properly utilize external tools. Accordingly, agent system 1 allows the AI ​​agent to perform tasks appropriately.

[0026] For example, as shown in Figure 5, the prompts of the inference instruction tool 12 contain abstracted inference steps. This configuration reduces the burden required to build the inference instruction tool 12 compared to a case where multiple specific inference steps are provided for each of the multiple specific examples. Furthermore, using abstracted inference steps improves the generalization performance of the inference instruction tool 12.

[0027] Various aspects of the invention derived from the embodiments described above are described below.

[0028] An information processing method according to one aspect of the invention is an information processing method in an agent system using artificial intelligence, wherein the agent system includes a reception step of receiving user input, a selection step of the agent system selecting one inference procedure suitable for the received input from a predetermined plurality of types of inference procedures, and a processing step of the agent system performing processing related to the received input according to a plurality of inference steps included in the one inference procedure.

[0029] In the information processing method of the above embodiment, the agent system may be configured to use a plurality of external tools to perform processing related to the received input, the plurality of inference procedures may be included in one of the plurality of external tools, and the description of the one external tool may state that the one external tool should be called first.

[0030] In the information processing method of the above embodiment, the agent system may use a large-scale language model as the artificial intelligence.

[0031] In the information processing method of the above embodiment, the user may be a occupant of a vehicle.

[0032] The present invention is not limited to the embodiments described above, and can be modified as appropriate without contradicting the gist or idea of ​​the invention as can be read from the claims and specification as a whole. Information processing methods involving such modifications are also included within the technical scope of the present invention. [Explanation of Symbols]

[0033] 1…Agent system, 2…Vehicle, 10…Server, 11…AI agent, 12…Inference instruction tool, 13…Calculation tool, 14…PoI search tool, 20…Terminal device

Claims

1. An information processing method in an agent system using artificial intelligence, The agent system includes a reception process for receiving user input, The agent system includes a selection step of selecting one inference procedure suitable for the received input from a predetermined set of multiple inference procedures, The agent system includes a processing step which performs processing related to the received input according to a plurality of inference steps included in the one inference procedure, Information processing methods including

2. The agent system is configured to use multiple external tools to perform processing related to the received input, The aforementioned multiple types of inference procedures are included in one of the aforementioned multiple external tools, The description for the aforementioned external tool states that the aforementioned external tool should be called first. The information processing method according to claim 1.

3. The agent system uses a large-scale language model as its artificial intelligence. The information processing method according to claim 1.

4. The aforementioned user is a occupant of the vehicle. The information processing method according to claim 1.

Images