The objective was to generate a whitepaper on FERC Order 2222 using Generative AI technology. It is widely believed that Generative AI technologies are capable of collating widespread information on any enquired / prompted topic and present it in choice of structure. This whitepaper validates the approach. The authors of this whitepaper have familiarity with FERC Order 2222 and structured the content generated by GenAI. The team captured their use of a variety of prompts to extract the correct content. Almost 90% of this whitepaper’s content is generated using Generative AI.Â
OpenAI responded to most of our queries on the topic of interest and smartly processed information from various sources to generate relevant context-based information. While it did not write the entire whitepaper, it was able to suggest some of the table of contents. We must change various prompts and train the model with correct inputs in order to receive the most relevant response. OpenAI was not able to generate images or videos for the given context as it is a Large Language Model (LLM) engine. OpenAI listed various components to be part of the architecture and suggested to build architecture views. Sometimes the responses were repeated even with a different set of input parameters. It is important to provide smart prompts to get an effective response from Generative AI. Finally, it was humans that validated the content as well as designed and edited the paper to develop the final product. It is also important to note that the real-implementation experience for some of the sections can be authored by humans only.
This article was co-authored with Sandeep Dayal.