Generative artificial intelligence (ai) to generate information related to a complex system

A system comprising one or more processors to:

• repeat generating a potential answer until the potential answer is determined to be responsive to one or more prompts, generating the potential answer to comprise:
• using one or more neural networks to generate output based at least on first information comprising the one or more prompts, the output to comprise at least one of a first answer, a request for second information, or a request to generate a tool, the potential answer to comprise the first answer if the output comprises the first answer;
• using the one or more neural networks to generate the tool based at least on the request to generate the tool if the output comprises the request to generate the tool, the potential answer to comprise results obtained by using the tool if the tool is generated and used to obtain the results;
• and
• determining whether the potential answer is responsive to the one or more prompts.

A method comprising:

• using one or more first neural networks to generate output based at least on a first prompt, the output to comprise at least one of a first answer, a request for particular information, or a request to build a tool to obtain data, a potential answer to comprise the first answer if the output comprises the first answer;
• using one or more second neural networks to generate the tool based at least on the request to build the tool if the output comprises the request to build the tool, the potential answer to comprise the data obtained by using the tool if the tool is generated and used to obtain the data;
• determining whether the potential answer is responsive to the first prompt;
• and
• using the one or more first neural networks to obtain additional information if the potential answer is determined not to be responsive to the first prompt.