Systems and methods for generating banks of age-stratified large language models

At least one computer-readable storage medium storing processor executable instructions that, when executed by at least one hardware processor, cause the at least one hardware processor to perform a method for generating a first bank of age-stratified large language models (LLMs) to simulate responses to prompts that would be provided by a first person in different life stages of their life, the method comprising:

• receiving, from at least one first datastore and via at least one communication network, first electronic data associated with the first person, the first electronic data received from the at least one first datastore corresponding to the different life stages of the first person, the different life stages comprising a first life stage and a second life stage of the first person;
• in response to receiving the first electronic data corresponding to the different life stages of the first person from the at least one first datastore, processing the first electronic data to generate first life stage data comprising first multiple data portions corresponding to the different life stages of the first person, the first multiple data portions comprising a first data portion corresponding to the first life stage and a second data portion corresponding to the second life stage;
• generating the first bank of age-stratified LLMs by using the first multiple data portions to further train a reference LLM, the first bank of age-stratified LLMs comprising a first age-stratified LLM corresponding to the first life stage and a second age-stratified LLM corresponding to the second life stage, the generating comprising:
• generating, for each particular data portion of the first multiple data portions corresponding to a particular life stage of the first person, a respective age-stratified LLM by further training, using the particular data portion, an instance of the reference LLM to generate simulated responses to questions that the first person would have provided in their particular life stage;
• storing the first bank of age-stratified trained LLMs comprising the first age-stratified LLM and the second age-stratified LLM;
• and
• using the first age-stratified LLM of the first bank of age-stratified LLMs corresponding to the first life stage of the first person and a second age-stratified LLM of a second bank of age-stratified LLMs corresponding to a second life stage of a second person to simulate a dialogue that the first person and the second person would have had in their respective life stages by:
• executing the first age-stratified LLM to generate a first simulated response to a question comprised in a first prompt that the first person would have provided in the first life stage of the first person;
• and
• executing the second age-stratified LLM to generate a second simulated response to the first simulated response that the second person would have provided in the second life stage of the second person.

A system for generating a first bank of age-stratified large language models (LLMs) to simulate responses to prompts that would be provided by a first person in different life stages of their life, the system comprising:

• at least one hardware processor;
• and
• at least one computer-readable storage medium storing processor executable instructions that, when executed by the at least one hardware processor, cause the at least one hardware processor to perform a method comprising:
• receiving, from at least one first datastore and via at least one communication network, first electronic data associated with the first person, the first electronic data received from the at least one first datastore corresponding to the different life stages of the first person, the different life stages comprising a first life stage and a second life stage of the first person;
• in response to receiving the first electronic data corresponding to the different life stages of the first person from the at least one first datastore, processing the first electronic data to generate first life stage data comprising first multiple data portions corresponding to the different life stages of the first person, the first multiple data portions comprising a first data portion corresponding to the first life stage and a second data portion corresponding to the second life stage;
• generating the first bank of age-stratified LLMs by using the first multiple data portions to further train a reference LLM, the first bank of age-stratified LLMs comprising a first age-stratified LLM corresponding to the first life stage and a second age-stratified LLM corresponding to the second life stage, the generating comprising:
• generating, for each particular data portion of the first multiple data portions corresponding to a particular life stage of the first person, a respective age-stratified LLM by further training, using the particular data portion, an instance of the reference LLM to generate simulated responses to questions that the first person would have provided in their particular life stage;
• storing the first bank of age-stratified trained LLMs comprising the first age-stratified LLM and the second age-stratified LLM;
• and
• using the first age-stratified LLM of the first bank of age-stratified LLMs corresponding to the first life stage of the first person and a second age-stratified LLM of a second bank of age-stratified LLMs corresponding to a second life stage of a second person to simulate a dialogue that the first person and the second person would have had in their respective life stages by:
• executing the first age-stratified LLM to generate a first simulated response to a question comprised in a first prompt that the first person would have provided in the first life stage of the first person;
• and
• executing the second age-stratified LLM to generate a second simulated response to the first simulated response that the second person would have provided in the second life stage of the second person.