The buzz around artificial intelligence (AI) in real estate now transcends basic chatbots and means a transformative era of structural change.

Property investors who engage with AI early stand to command the largest slice of those gains, says Ian McGuinness in Knight Frank’s latest Wealth Report.

“In the property sector, rapidly improving technology offers a once-in-a-generation opportunity to make money.”

Global real estate company Knight Frank’s Wealth Report section on AI shows the existing spotlight is on improving productivity in knowledge-centric roles – for example, streamlining mortgage assessments and valuations, and identifying sites for development.

But for the luxury residential sector, dominated as it is by heightened demand for personalised experiences, AI looks set to offer sophistication, customisation and efficiency, providing insights into client behaviour, preferred locations, favoured architects and developers and lifestyle preferences.

Until now successful developers relied heavily on who they knew, which in turn dictated how they found the best sites and accessed the finance to develop them.

“Success in this industry has always been about insider knowledge. That’s why it’s so exciting that we are at the beginning of a revolution in AI – one that will enable us to make sense of quantities of information that would have seemed impossible only a few years ago,” says McGuinness, head of research analytics.

Expanding AI-driven data processing capabilities will result in the largest improvement to productivity since the arrival of the internet, saving the average worker more than 100 hours a year, according to Google. In the UK alone Google estimates that AI-powered innovation is expected to create $NZ833 billion in economic value by 2030. GDP is expected to be more than 10 per cent higher over the same period than it otherwise would have been, according to PwC.

The explosion is here

The explosion in popularity of large language models such as Open AI’s ChatGPT can make it appear as though AI only arrived last year, but the natural language processing (NLP) technology underpinning these chatbots has been around for several years.

The beauty of ChatGPT is its simple interface that enabled hundreds of millions of people to get their first taste of what AI might offer. NLP enables people or companies to extract value from huge amounts of unstructured text.

Until now, getting that information has been frustrating as it involves manually moving data from a PDF into a spreadsheet so it can be manipulated and mined for insights.

AI automates that process on an enormous scale. These models can put words into context and draw inferences in a manner that the world is only just beginning to understand.

The future

Knight Frank began experimenting with NLP four years ago by turning it loose on planning databases.

It immediately became apparent there were troves of information in the thickets of commentary added by planners that would be impossible to compile manually.

“We were able to track how businesses across the country were adapting to the Covid-19 pandemic by changing the use of properties, for example,” McGuinness says.

It was immediately clear NLP could be used most profitably to find sites, particularly when paired with other techniques.

Over a coffee, AI now allows McGuinness to take a client on a walk through prime central London and point out every available development site and explain their positioning relative to lifestyle elements such as the highest performing schools, fine dining spots and private members’ clubs. He could also point out how many wealthy individuals live nearby, their property preferences, plus how much they have to spend.

Seniors’ housing

These technologies get more interesting when they are applied to complex problems, particularly during periods of elevated borrowing costs.

The financial viability of retrofitting commercial buildings varies depending on underlying values or rents. By feeding its models with the relevant data, Knight Frank can produce accurate contour lines over a vast area that display exactly where investing is most likely to generate the best returns.

Seniors’ housing is another sector for which development viability turns on dozens of factors. People over 60 tend to buy property only within three miles of where they live, so any feasible development site must have the depth of market nearby.

“Developers come to us for highly localised answers as to where it is feasible to start digging, but at the touch of a button we can apply our criteria to the entire country, at times providing hundreds of suitable sites,” McGuinness says.

One of the many examples of scalability was Knight Frank’s study of car park land for officials at a local council department, who wanted to find out whether any of its owned car parks could be put to more valuable use, such as retail or residential.

Knight Frank studied more than 30,000 car parks and found that, while they were well served by public transport, almost 70 per cent didn’t appear to support a retail centre. Considering land values and redevelopment potential, it identified enough land to support more than 100,000 new homes, all while maintaining car parking services vital in supporting high streets.

‘Lightning in a bottle’

Data is rapidly proliferating, and how real estate investors manage so much of it is now among their biggest challenges.

McGuinness says AI will help capture “lightning in a bottle” opportunities, because only its computational power can fully resolve and scale the myriad physical, socioeconomic and environmental conditions underpinning success.

“We view AI as a chain of processes that includes language processing but also spans machine vision, classification and learning techniques that enable us to address data gaps and make clear-eyed predictions.

“The abundance of information fuels the potential for a generation of property investors to achieve new heights – if they can bring fresh perspectives and discern the signals within the noise.

“Returns will increasingly correlate with the quality of the data and interpretation they have access to. Success is still about ‘insider knowledge’, but that term no longer means what it used to,” he says.