In the last decade, we’ve witnessed a surge of built environment startups. When we look at the data, the numbers tell a compelling story. Prior to 2010, this sector saw fewer than 100 startups founded annually. Fast forward around 2018, and the landscape reached a remarkable peak with an impressive 700 startups emerging each year.
This increase in ConTech on the market tracks with broader advancements in technology, including the release of the first iPad in 2010. Technology enabled construction companies and real estate firms alike to embrace a digital transformation.
Today, we are at another notable inflection point with the recent advancements in Large Language Models (LLMs) and the introduction of generative Artificial Intelligence (AI) tools. Applications of these technologies introduce the potential to streamline industry processes and accomplish goals that were deemed impractical or prohibitively expensive not long ago.
At Building Ventures, we have witnessed a surge in AI solutions tailored specifically for the built environment and expect to see this trend continue. As more AI-backed startups launch in our sector, it will only become more critical to accurately assess these technologies. In this article, we will share a few key methods we use for evaluating AI-powered technologies to identify promising solutions to create a better built world.
Let’s begin by addressing a fundamental aspect of AI: training data. Since 2010, the construction industry has embraced a digital transformation, including leveraging the cloud for data storage. There have been a number of motivating factors for General Contractors (GCs) in particular, including the increasing complexity of projects, growing cost of materials, and widespread labor shortages. However, GCs have recently grown more concerned about data security and privacy.
Given this landscape, architecture, engineering, and construction experts have become increasingly cautious about the teams they collaborate with and how their data is shared. All founders in our sector need to be aware of this growing skepticism, but it’s especially important for AI-powered technology. These startups require data collection to train their models, which necessitates a supply of large data sets and a proprietary claim. Because the stakes are higher, these founders need to consider how they collect data and how it will be used. This is a pivotal consideration, as unique access to specific data can serve as a key differentiator in a competitive landscape where multiple entrepreneurs are pursuing similar solutions.
This competitive landscape is another crucial aspect we consider when evaluating AI-powered technologies for the built environment. We ask the same fundamental question of all startups: What specific problem does the solution address? Because the current environment is saturated with AI-backed, it’s important to discern whether the proposed innovation genuinely resolves a pain point or merely automates complexity.
The construction industry is known for its thin margins. This makes resistance to adoption a significant challenge, unless there is an undeniable opportunity for project improvement. It’s imperative to recognize that, although AI is amazing technology, our investment focus is on solutions that contribute to creating a better built world. That means we must prioritize tangible outcomes over merely showcasing technological capabilities — especially in an industry with such serious risks.
Another pivotal consideration when thinking about generative AI is the question of responsibility and risk associated with the outcomes produced by the solution. In the built environment, we are dealing with tangible outcomes subject to code compliance, jurisdictional legislation, owner stipulations, and potential litigation in the event of failure, to name a few factors.
Any lapses in asset performance or deviations from compliance standards can lead to serious consequences, including legal actions or, in the worst-case scenario, the loss of lives. This underscores the gravity of accountability and risk management in the application of generative AI within the built environment, and we keep this top-of-mind when evaluating a technology.
Additionally, it is crucial to assess how the solution is keeping pace with forthcoming regulations concerning AI. Given that much of the regulatory landscape in this domain is still evolving, founders need to stay proactive in tracking and, in some cases, even anticipating these changes. Notably, instances of litigation related to intellectual property, such as the high-profile case involving GitHub’s Copilot, highlight the legal intricacies in this emerging field.
Relatedly, the question of unit economics arises. How will the teams sustain growth when facing an exponential user base, particularly if they rely on existing models that typically start with a free usage tier but shift to charging as usage spikes? AI models are computationally intensive, posing scalability challenges that need careful consideration as usage scales up. Founders need to address these challenges in their plans for sustained growth beyond the initial stages.
Our final key consideration is discerning the distinction between industry-specific solutions and mainstream counterparts. In a landscape where many companies are already engaged in AI experiments, often with major players like Microsoft, Alphabet, and Autodesk, startup solutions must showcase genuine innovation in addressing the specific challenges they aim to solve. Demonstrating unique problem-solving approaches becomes a crucial factor in standing out in a market saturated with AI initiatives.
For those startups able to differentiate their technology, there is tremendous potential for growth within our industry. In fact, we contend startups focusing on developing AI applications for specific vertical markets stand a higher chance of success.
As we have shared throughout this article, we are genuinely enthusiastic and optimistic about the impact generative AI will have on the built environment. It is our responsibility as investors to ensure that the solutions we back are poised for exponential growth, have the power to transform the industry, and will be developed responsibly. Ultimately, our overarching focus remains steadfast: contributing to the creation of a better-built world.
