AI and the Future of Real Estate
Recent advancements in generative AI have the potential to remake how real estate investments are sourced, designed, managed, and more. Today we'll separate the promising applications from the buzz.
Thesis Driven dives deep into emerging themes and real estate operating models by featuring operators executing on each theme. This week's letter is a guest post from Hewsan Pang, Founder Fellow at On Deck and former Managing Director at Common. It analyzes the impacts of generative AI on the real estate sector.
Over the past few months, AI has captured the internet’s imagination. Recently released tools—including ChatGPT-3 (text generation), DALL-E, and Stable Diffusion (both image generation)—seem almost magical in how they can easily create human-like content such as poetry, college essays, code, texts-to-images, and more. Everyone is now looking for ways to use this new generation of AI to his or her advantage, and real estate is no exception. Today we will focus on the AI tools that exist today, those in the works, and what's possible in the future that could transform real estate investment and development as we know it.
Earlier generations of AI offered predictive modeling by identifying relevant patterns in data which then could be used for better decision-making. A great example of this technology at work in real estate is revenue management software, such as RealPage's YieldStar (now called AI Revenue Management) for multifamily or Duetto for hospitality. These software tools use market data to forecast supply and demand dynamics, suggesting daily price changes for available units.
What makes the new generation of AI more powerful—and more awe-inspiring—is that it's been trained on incredibly large datasets to not only identify patterns in the data, but generate new data on demand. We've already seen glimpses of this technology at work. For example, Gmail predicts the ending of our sentences as we write them and Grammarly not only checks our grammar but also suggests text replacements. However, these tools have been fairly limited in scope. With OpenAI's public launch of GPT-3 with API endpoints and its new Foundry developer platform, we are on the cusp of witnessing the transformative potential of AI across industries.
While some offer wild predictions for what AI can achieve, we will dive into four specific use cases within the real estate industry that are possible based on the AI models that are available today:
Marketing and customer service
Generative design
Copilot applications
Structuring unstructured data
In addition to discussing these use cases, we’ll preview some of the challenges and emerging opportunities that real estate operators face as they push to adopt AI in the wild.
 | A guest post by
|