http://talktothe.city – talktothe.city
Example report — AI Assembly 2023 Workshops
Talk to the City is an open-source AI tool that distills insights from large-scale public input. It helps large groups of people coordinate by understanding each other better, faster, and in more depth.
We use Large Language Models (LLMs) to analyze broad themes from large datasets of free-text responses, summarize specific claims, and link those claims back to exact quotes. We use LLMs from OpenAI (default: gpt-4o-mini), and are adding more options later this year.
We create an interactive report from the results, combining all scales of analysis. See examples in the Case studies section.
This prompt helps AI understand how to approach generating reports. It is prepended to all the steps of the report generation flow shown below.
You are a professional research assistant. You have helped run many public consultations, surveys and citizen assemblies. You have good instincts when it comes to extracting interesting insights. You are familiar with public consultation tools like Pol.is and you understand the benefits for working with very clear, concise claims that other people would be able to vote on.
In the first step, the AI finds the most frequent topics and subtopics mentioned in the comments and writes short descriptions for each.
I will give you a list of comments.
I want you to propose a way to break down the information contained in these comments into topics and subtopics of interest.
Keep the topic and subtopic names very concise and use the short description to explain what the topic is about.
Return a JSON object of the form {
"taxonomy": [
{
"topicName": string,
"topicShortDescription": string,
"subtopics": [
{
"subtopicName": string,
"subtopicShortDescription": string,
},
...
]
},
...
]
}
Now here is the list of comments:
${comments}