In an era where Artificial Intelligence (AI) is rapidly reshaping industries, governments are no exception. The advent of Generative AI, capable of creating new content like text or images by learning from existing data, offers tantalizing possibilities for enhancing public governance.
A few months ago, I outlined a modest proposal for a customized AI engine trained on public documents that could provide authoritative, interactive responses to a myriad of questions from citizens about city planning, regulations, and other civic matters. Imagine a city where residents get real-time, accurate answers from a state’s website as if conversing with an expert from the relevant department. By leveraging models like ChatGPT, governments could dramatically improve data access, transparency, and efficiency, paving the way for more informed and engaged communities.
The City Planning Chatbot Experiment
I collaborated with Code Studio, an urban design and zoning firm based in Austin, TX, to develop this prototype City Planning Chatbot. We used publicly available information from the City and County of Denver to experiment with this concept.
Disclosure: This effort is not endorsed or supported by the City or County of Denver. I used publicly available information on their website to develop this prototype.
Here are some examples of questions I posed to the City Planning Chatbot about zoning in the City of Denver.
Question: What is the allowable height of a fence in a residential zone district?
Chatbot Response: In a residential zone district, the maximum height of a fence when forward of any primary street facing primary structure facade is 4 feet. However, when the fence is located at or behind any primary street facing primary structure facade, the maximum height allowed is 6 feet. This is according to Section 10.5.5.2 of the Denver Zoning Code.