ChatGPT Use Cases: Example #1

"Well-Functioning Urban Environment"

ChatGPT4 generated prompt: Create an image inspired by Kevin Lynch's urban analysis diagrams …

February 2023

One of my first experiments using ChatGPT was way back in February 2023. About six months earlier I had completed a piece of work for the Far North District Council to undertake an independent “on the ground” assessment of “functionality and form” of the urban or peri-urban areas of Kerikeri-Waipapa.

Together with Tim King at Common Ground Studio we developed an evaluation and scoring system that blended evaluation criteria from the Lynch Analysis methodology with those of the NZ Urban Design Protocol, National Policy Statement on Urban Development, the Scottish Government's Place Standard, and the metrics used by the OECD and the NZ Treasury to measure Wellbeing. For our report, these were all aggregated and then distilled into 10 categories that we were able to measure, spatially model using GIS, and report on:

1. The variety of homes

  • House Size

  • Residential Density

  • Variety

  • Urban Intensification

  • Flexible Planning Options

  • Affordability

  • Mixed Use (Biz in Res zone)

  • Servicing

 2. Variety of sites for different business sectors

  • Variety of Business Activities

  • Mixed Use (Res in Biz zone)

 3. Competitive land markets

 4. Accessibility

  • Network Connectivity - Road

  • Network Connectivity - OS

  • Range of Modes

  • Legibile and Coherent

  • Connected Destinations

  • Carparking

  • Arterial Access

  • Social Infrastructure

 5. Open-space

  • Open-space Amenity

  • Parks/Open-space Proximity

  • Esplanade Access

  • Natural Character

 6. Urban Design

  • Legibility - Urban Form

  • Strong Urban Identity

  • Sense of Place, Memorability

 7. Contribution to climate change resilience

  • Mitigation

  • Environmental Protection

 8. Māori cultural norms and traditions

  • Unique Maori Identity

  • Heritage Protection

 9. Community

  • Dynamic Urban Life

  • New Tech

  • Arts

 10. Collaboration

  • Co-ordinated Actions

  • Transport Corridor Impacts

  • Tangata Whenua Relationships

Testing knowledge by pleading ignorant

I wanted to know how I could have used ChatGPT for the project design phase had it been available at the time. To test its knowledge and capabilities, I used a line of enquiry based on open questions to see where it would take me, starting with:

ChatGPT3.5 was familiar with the concept. Without providing context (such as referencing the NPS-UD) I asked:

Next, I introduced the NZ Urban Design Protocol with a prompt asking ChatGPT to 'compare and contrast' with the Lynch Analysis.

Over the next series of prompts, I asked if either analysis technique could be used to arrive at a score (say 1-5) for how well-functioning an urban area is, and what data sources could be used to calculate a score. All of ChatGPT’s answers were consistent with the approach that we had already used for the project we had completed six months earlier. For our project, we also drew from the Scottish Government’s Place Standard. ChatGPT could see how that tool might be used for quantitative analysis of '“well-functioning” urban environments:

IDEATION (AGGREGATING METHODS)

For our project, we had aggregated a list of criteria that incorporated all three evaluation methods. I asked ChatGPT to do the same:

 "Give me a list of criteria that aggregates the Lynch Analysis, NZ Urban Design Protocol, and the Place Standard.”

Here is a list of criteria that aggregates the Lynch Analysis, the New Zealand Urban Design Protocol (NZUDP), and the Scottish Government's Place Standard:

 1. Accessibility: The extent to which the built environment is accessible and supports mobility for all users, including people with disabilities, children, and elderly people.

 2. Connectivity: The quality and legibility of the transportation network and the connections between different parts of the city.

 3. Land use mix: The diversity and balance of land uses in the built environment, including residential, commercial, and public spaces.

 4. Sustainability: The extent to which the built environment supports environmental sustainability, including energy efficiency, resource management, and the protection of natural resources.

 5. Safety: The extent to which the built environment supports personal safety, including measures to prevent crime, promote pedestrian safety, and ensure fire safety.

 6. Imageability: The visual and aesthetic quality of the built environment, including the design of buildings, streetscapes, and public spaces.

 7. Social well-being: The extent to which the built environment supports social well-being, including the provision of public spaces for social interaction and recreation.

 8. Adaptability: The extent to which the built environment can adapt to changing needs and conditions, including the provision of flexible and resilient infrastructure.

 9. Distinctiveness: The extent to which the built environment has a unique character and sense of place, including the presence of landmarks and other distinctive features.

 10. Economic viability: The extent to which the built environment supports economic viability, including the provision of employment opportunities and the promotion of economic growth.

 11. Health and wellness: The extent to which the built environment supports health and wellness, including the provision of parks and other green spaces for recreation and relaxation.

 12. Cultural heritage: The extent to which the built environment supports the preservation and promotion of cultural heritage, including historic buildings, monuments, and landmarks.

 13. Quality of life: The overall quality of life in the built environment, including the provision of services, amenities, and opportunities for personal growth and development.

 14. Community involvement: The extent to which the built environment supports community involvement and engagement, including opportunities for public participation in decision-making and the promotion of community cohesion.

 15. Responsiveness to context: The extent to which the built environment responds to its context, including the physical, social, and cultural context of the area.

 16. Efficient use of resources: The extent to which the built environment supports efficient use of resources, including the use of renewable energy sources, water conservation, and waste reduction.

 This list of criteria aggregates the key elements and design principles of the Lynch Analysis, the NZUDP, and the Place Standard, and provides a comprehensive framework for evaluating the quality and functionality of an urban environment. The specific criteria used in a given evaluation would depend on the goals and objectives of the analysis, as well as the context of the urban environment being evaluated.

VALIDATING THE OUTPUTS

Some of my initial thinking on this project was to align our evaluation with metrics that were measured and reported on at national and international levels. Having a consistent and replicable tool that was used for evaluating all urban environments would enable the data to be aggregated and fed into these higher-level reporting frameworks. Well, that was the theory anyway, and is why I provided for this scenario as part of my project design. In practice that scenario has its complications, which ChatGPT was able to clearly describe:

My Conclusion on this use case

How well did ChatGPT3.5 do? Well, in just a few minutes I had a list of criteria that could be used to measure how well an urban environment function. It incorporated all the criteria that needed to be included and is pretty consistent with the criteria that Tim King and I put together over several days.

As a tool for ideation, comparing, contrasting, and alignment of different methodologies (even very different ones), ChatGPT can save literally hours of time!