Evolution 4.0: The AI revolution in commercial real estate

In many ways, chess was the perfect place to forge a new, more complex form of artificial intelligence (AI). Simultaneously rigidly codified and fiendishly intricate, it is a game that rewards heightened understanding of strategy, probability and risk. In 1997, having lost the year before, an IBM computer named Deep Blue™ beat reigning world-champion Gary Kasparov by two games to one over a six-game series. It marked a watershed moment in the evolution of humankind’s relationship with machines. Deep Blue™ is now enjoying a hard-earned retirement in Washington DC’s Smithsonian museum. This is perhaps for the best – were it capable of emotions, it might well be feeling much older than its 23 years of age. AI has come a long way since the late 1990s.  

These days, with a few exceptions, when we talk about AI, we mean machine learning. Strictly speaking, machine learning is a subset or application of AI. Deep Blue was fed volumes of specifically selected data with the ultimate goal of completing a single, albeit ambitious, task: defeating a chess grandmaster at his own game. However, machine learning is far broader. It is about feeding nonspecific data into a computer programme without a narrowly defined end goal. Instead, the objective is a general refinement of the programme’s ability to use information to draw conclusions. In a recent World Built Environment Forum webinar, Carlos Alvarez Ramallo, Industry Manager for Finance and Banking at Google in Spain, described this as engaging the machine in a pseudo-human behavioural pursuit. Ultimately, it is hoped that machine learning will be the tool with which we unscramble the puzzles that have stumped even the most capable human minds.

We are, in reality, only now scratching the surface of the full range of use cases for machine learning. It is particularly good at probability assessments, being dispassionate about the value of data and uninfluenced by the lived experiences that so often cloud human judgement. But it is capable of so much more. By aggregating declared preferences, it can improve customer experience through personalisation. This capability, for example, enables real estate agents to better identify properties likely to interest prospective home buyers. It could revolutionise property valuations by introducing for consideration lesser-used data relating to local demographic trends, density rates and local public transport service quality. This will furthermore enable real estate agents to furnish clients with a much wider range of pertinent information related to the liveability of a given neighbourhood or district. Analysis of construction delays and cost overruns can be used to forecast the previously unforeseeable and streamline a sector traditionally hamstrung by inefficiencies and waste. Similarly, analysis of use trends can be harnessed in commercial real estate to improve how space and energy are deployed – both in individual assets and across portfolios. Truly intuitive buildings are likely to become highly valuable in a post-COVID-19 world where remote work will be common and maximum workplace occupancy correspondingly rare.

Machine learning also will be a key driver of organisational change and ask a great deal of senior leaders. Implementation is likely to require significant organisational and structural change. To embed complex AI successfully into their daily operations, businesses must prioritise the development of foundational technological capabilities. These include IT infrastructure and cybersecurity systems alongside data management and governance processes. But, arguably more important still is the role that humans play.

When it comes to adoption of AI, the principal cause of failure is the lack of human insight and expertise. In last month’s Evolution 4.0 column, I stressed the primacy of people in digital transformation strategies. I do so again here. Human capital is among the most critical components in mitigating algorithmic bias.

Collaboration will be crucial for businesses seeking to make machine learning a key component of their working model. Responsibilities for data management, analytics, IT infrastructure, and systems development, as well as business and operational expertise, must be closely joined up. Of equal import will be the establishment of clear ethical, legal, reputational and financial risk management procedures. Many organisations with mature AI practices are now establishing policies for the governance of key technologies. Among these will be clear protocols for explaining how their algorithms deliver results. Clients may need some encouragement to embrace the algorithm – at least initially. Demonstrating comprehensive and transparent accountability frameworks will be integral in the battle to win their hearts and minds.

Responsibility for all of this will ultimately fall to the C-suite. Without a strong sense of purpose, transformational change is doomed to fail – be it through poverty of ambition or lack of direction. Up until now, responsibility for managing such transitions has tended to fall to CEOs, COOs and CIOs. Over the coming years, we can expect to see an expansion of chief officer roles. Moving forward, serious businesses will need a Chief Data Officer and Chief Analytics Officer.

The AI revolution is changing everything about analytics in commercial real estate and adjacent sectors. Human talent must continue to actively engage with AI to find relationships in the data, create visualizations, aid in storytelling and share findings within each enterprise and across these industries.