Exemplar: Using AI to Enhance Learning at Florida International University
Disclosure: Florida International University is an HP customer and has received loaned and/or donated equipment from HP.
John Stuart, a professor at Florida International University (FIU), had wanted to teach a class with AI for five years, but it wasn't until the meteoric rise of AI this year that he found a mechanism to bring it into his classroom. In one of his graduate architectural design courses, "AI-Assisted Regenerative Strategies for 2070: Future-Oriented Architectural Concepts and Climate-Adaptive Historic Preservation in Miami Beach," he's having students learn about and plan the future of historic buildings in Miami Beach. Each student in the course will be assigned a building in a designated historic area, where the buildings are vulnerable due to rising sea levels and increasing storm intensities. Stuart and his students are working with Giri Narasimhan, a data science professor at FIU, and his graduate students in computer science to train AI to assist in interpreting old, poorly scanned city-provided building cards (see figure 6).
The City of Miami Beach has thousands of these cards that come in different shapes and sizes, with handwritten notes and marks added to most. The students are working with the data scientist to collect the training data, with the goal of one day using AI to parse and interpret the information on the cards. Such assistance would make it much easier for students and, in the future, city staff members and property owners to gather background information on these buildings, helping them identify potential commonalities of date and type of construction, the architect and design, and contractors. With this information, they will be able to understand their building in relation to others in a smarter way. If a contractor helped build two or more buildings around the same time, they would likely be made with similar materials and, therefore, encounter similar problems. With this information, students will understand patterns of aging in buildings and could uncover ways to preserve their individual buildings against the issues they will potentially face in the coming 50 years.
In addition using AI to help understand the past about these buildings, students will also use AI to help them create their future "clients" for their buildings. Students will ask ChatGPT Plus with GPT-4 to speculate—on the basis of the history of the building they are studying and the people who have lived there—about who will be living there 50 years from now. The AI then jumps in and creates a series of future clients living in the building for the students to imagine how they might meet their needs by modifying and improving the building.
Example Prompt and Output from ChatGPT Plus with GPT-4:
PROMPT: Speculate on the future of 1600 Meridian Avenue, Miami Beach, Florida, in the year 2070. Imagine that the building has 15 two-bedroom apartments and 8 one-bedroom apartments with families living in each. Speculate on how the building is very sustainable and the people in it are dealing creatively through a variety of strategies with sea-level rise of about four feet. Speculate on the names and possible histories of the families living there. Speculate on where they might have been born and what their jobs/passions might be. Imagine a scenario in which you know their names and relationships to one another. And imagine brief histories and backgrounds for those living in the building and what kinds of things you imagine they might be doing in 2070.
Response from ChatGPT Plus with GPT-4
I'm glad to create a speculative scenario! Here are some fictional families living at 1600 Meridian Avenue in Miami Beach, Florida, in the year 2070:
The Martinez Family: Victor, an environmental engineer, Maria, a sustainable fashion designer, and their twin sons, Luca and Diego. Victor was born in Cuba but immigrated as a child, while Maria is a third-generation Miami native. Both work from home, and their apartment's balcony has been transformed into a hydroponic garden. They are known for organizing eco-fashion events.
The Stein Family: Eleanor, a marine biologist, and Ben, a VR educator, live here with their daughter, Lily. Ben is a New Yorker who moved to Florida for his love of scuba diving, and Eleanor was born in Orlando, Florida. Lily, a budding marine enthusiast, often helps her mom with research. They host virtual educational tours of Miami's aquatic life.
The Patel Family: Ravi, a climate lawyer, and his wife Meera, an AI therapist, and their two children, Anya and Ishaan. They were born in India but have been living in Miami for a decade. Their home is full of smart tech, and they enjoy hosting AI and law-related discussions.
The Jackson Family: Olivia, a local artist, and her son, Nate. Olivia was born and raised in Miami, and her art often reflects the transformed cityscape. They organize community art exhibits to raise awareness about climate change.
The Fujioka Family: Hiroshi, a software engineer, and Akiko, an urban planner specializing in flood-resistant cities, live with their son, Ken. Both were born in Japan but have been living in Miami for 20 years. They organize annual architecture and city planning workshops.
The Abdul-Rashid Family: Hamza, a solar-panel engineer, and Fatima, a telemedicine doctor, along with their daughters, Laila and Aisha. Born in Egypt, they moved to Miami a decade ago. They conduct workshops on renewable energy and health care technology.
The building itself would have a variety of strategies to deal with sea-level rise, such as stilt architecture, waterproof materials, rainwater collection systems, and flood barriers. Residents would likely use renewable energy, including solar and wind power. They might have community gardens, vertical farms, or even aquaponic systems to grow their food. They would likely rely on technology to monitor weather patterns and potential flooding.
The residents' bonds are strong as they collectively work toward a sustainable future. Each family brings unique skills and experiences to the table, fostering a vibrant community that's adapted to their changed environment. They share a resilient spirit, living proof that life not only survives but thrives in the face of adversity.
For their final projects, students will then use Midjourney, a text-to-image generative AI system, and their architectural skills to create visual representations of the historic preservation of their resilient buildings. The students ultimately present their projects to city staff and elected officials to inspire enlightened conversations in the community about effective policies to support the future of smart and sustainable historic cities.
Stuart shared additional lessons from this project:
- Engaging new technologies with a positive mindset can help impact conversations around the technology and the potential fears that surround it.
- Institutions may need to bring in outside help when it comes to AI to help faculty understand its potential for their classes.
- When looking to test new technologies such as generative AI, focus on how its use can serve students.