Top Story
ULI Poland September 2023 newsletter
Stay up to date with the latest ULI Poland news and upcoming events!
September 20, 2023
Real estate, land-use and urban planning professionals seem to be more eager than ever to harness the power of artificial intelligence in their everyday work. Will the AI-enhanced property solutions be ready for their high hopes and complex market-specific needs?
The fully packed KinoGram auditorium in Warsaw’s Fabryka Norblina on September 12th left no doubt about what makes the industry tick in 2023. Whether generative, or predictive, artificial intelligence capabilities are at the heart of everyone’s interest. Its ongoing, wide-spread commercialisation around the world sparks curiosity, creativity, concerns, and spending sprees, also amongst the Polish market participants.
“There is plenty of interesting opinions and discussions swirling around about the ways in AI can change the ways in we work and live,” Soren Olsen, chair of ULI Poland, pointed out when welcoming the guests of the completely sold-out ULI Poland Places+Spaces “AI – What does it mean to the real estate sector?” event, which gathered both investors, developers, brokers, consultants, architects, and lawyers. While eying AI-enhanced business opportunities and risks – they wonder how to best navigate the still early days of AI to truly gain advantage and resilience, and not marketing gimmicks, in the years to come.
Reimagined future under supervision
More customized and accurate recommendations, enhanced price predictions, searches, stronger client relationships and portfolio optimizations, faster and more transparent buying processes, improved property visualizations, and urban modelling practices – these are just some of the property world’s fast-growing expectations towards AI opportunities.
Yet, it must not be forgotten that they should be thoroughly confronted with the less appealing side of the coin for the complete business assessment, namely threats and reservations, especially about potential manipulations, misinformation, misrepresentations, bias or the co-called hallucinations, prof. Aleksandra Przegalińska, an associate professor at Kozminski University and AI researcher affiliated with Harvard University, explained in her opening speech, partly presented by her AI-powered avatar, much to the pleasant surprise of attendees.
“If the training data contains biases, generated content will likely amplify those biases leading to outputs that may be discriminatory or unfair. The AI’s potential is vast, but as with any transformative technology – it brings along ethical and practical challenges that we must address with care and foresight. (…) These are just statistical models – they are blind to intention. They don’t distinguish between a lie and truth. They don’t know when they are producing something that is completely fictive and non-existent and human supervision is crucial,” she elaborated.
Does it mean that the labour market will not see any major AI-driven shakeups anytime soon as feared? Not necessarily. The market might drastically change, but not in the ways most expect: AI is likely to free up human workers’ time resources spent on repetitive or manual tasks once the technology’s advanced assistive features develop, rather than to replace them. But for this to succeed, the machine learning algorithms would need an enormous assistance too: the right prompting from human engineers, which could create a new job market demand.
The experiments prof. Aleksandra Przegalińska has been conducting with her teams on marketers in the field of collaborative AI, where bots are designed to work out joint conclusions and answers together with humans, and not instead of them, show some promising results. A 57% increase in job satisfaction levels, and an app. 40% boost in job efficiency in both routine and creative task categories, along with higher performance quality and speed, were observed.
Something new, something old
But due to the local market specifications, the Polish AI aficionados might encounter a bumpy road ahead, before the IT revolution gets into the full swing in the country. Poland’s irrevocably ageing building stock would be one of the first barriers to the effective introduction of any AI-driven property innovations at a large scale. “If a building features a 20-year BMS then we would probably be looking at investing into a new one first,” said Renata Hartle, technology solutions director at Colliers, who shared her presentation highlighting, among others, a significant digitalization gap observed in app. 50% of the Polish real estate assets. “This year, I audited over 25 buildings, and most of them had disconnected systems, still relied on manual processes and work, even with such tasks as the meter read-outs,” she said.
“I would encourage everyone to really look broader on how to prepare the building systems for further data management and integration,” she added. She supported her point with a showcase of some excellent Polish use cases, where AI equally enhances day-to-day work of tenants, managers, and developers. “Applying AI to the access control system can let us predict how many people would come to a building on any given day, and how we should respectively regulate temperatures or manage lighting system to save energy,” she explained. AI solutions are believed to have the potential to help reduce 10-20% of energy costs and up to 10% of GHG emissions.
Speed of change
While the pandemic years might have accelerated digitalization efforts in the property industry in Poland, the urgency to prioritize IT solutions didn’t last for long. As pointed out by Tomasz Ogrodzki, founder, and CEO of REDD Group, who cited in his presentation the ULI’s 2023 Emerging Trends report, digital transformation is one of today’s least pressing issues for the Polish property industry. “We know that inflation and economy are more important,” he said. This short-sighted perspective will likely continue to be a major roadblock standing in the way of companies trying to tap the full business potential of AI transformation.
Its promise, which doubles every three months, will certainly not wait for the laggards. “Physicality is the biggest strength and the weakness of the industry at the same time. It makes the change [towards AI] trickier because of the high cost of digitalization. And, on the other hand, it also makes physical adaptation impossible or very challenging. We could say that the office crisis is the best example for that: we do not adapt as fast as we would like,” Tomasz Ogrodzki noticed. Sped up revolution would equally require more and better data, he added, before presenting the ways in which AI is applied by his team to support real estate valuations.
Worth all the money?
In the following expert discussion panel, moderated by Tomasz Aleszczyk, an ULI Poland executive committee member, guests faced a round of questions from the host, and audience members. The kept coming like waves: What do companies spend on AI rollouts in Poland? How and when would it pay off? How much quicker would property transactions become with the help of AI? What does the future hold for AI-powered workforce
“Gradually, I think it could be doable to reduce the time of investment process by a half,” said Piotr Trzcinski, head of Poland at Savills Investment Management. “There are of course some aspects which will remain out of control of the parties involved. Central repositories or databases run by the state regulators, and granting access to the AI tools could help out. They could greatly speed up the transaction process. (…) Just to harness your own data, it goes well into tens of thousands of dollars. Let alone, trying to collect data from the market,” he added. And reminded that AI “does not absolve us from doing our own research and validation.”
Meanwhile, according to Michał Jaskólski, co-founder, vice president and CPO at Morizon-Gratka Group, there could be two ways of looking at the AI spending commitments companies try to make. “You can take your R&D budget and decide to spend as much of it as you can on AI. The other approach is a boarder analysis: ask yourself how much differently you would build your company and processes with AI, if you were to build them from scratch right now, so you could outcompete yourself and your competition,” he told the event’s participants. In his eyes, the sacrifice of app. 50% of R&D budget seems “just right.”
But at the end, although AI investments are now made mainly with meaningful cost reduction and revenue increases in mind, projects which serve a higher purpose of, say, improving consumer protection and antitrust policies in respect to AI, and lead to better corporate controls over the technology, should not go unnoticed either, Michał Jaskólski added. To strengthen his point, he mentioned Zillow’s failed algorithmic house-flipping business, whose executives allegedly had convinced developers to alter the risk precautions for higher gains.
Human touch
As it turns out, money will not get you everything in AI, too. “If you look at the quality of data that that we require in the investment business, we do need the right context, not just raw data. For example, understanding how the yields were construed, whether there was a vendor loan involved. We need the underlying information, with which AI will now have difficulties on its own. Only human-to-human interaction can produce this information,” said Piotr Trzcinski.
The paramount importance of social impacts and interactions was also touched upon by Renata Hartle: “When I think of my daughter, and how she’s going to grow up in the world full of AI, I’m worried about the future generations, about people not being able to differentiate deep fakes and bots from reality, for example, or technologies destroying their mental health. AI could give more power to whatever and whoever is already destroying our world. (…) We need to understand that we are just at the beginning. We have no idea how far we can get.” she summed up.
Greater responsibilities Â
Audience members also shared their reflections and AI experiences with the conference, providing an even more wholesome it pictures of where the Polish market and AI stand today. “I think development is an art. A machine is never going to replace an artist. You can ask it to do it, but I don’t think there is going to be a satisfactory result there,” said an audience member. Another, a lawyer, admitted that the tool he had used had made “horrible mistakes when it comes to accuracy.” As the models employed by AI still lack sufficient expert data, it is very difficult to see it taking over the legal profession any time soon. “However, at the same time, I believe there is a huge potential there: to me the law is about data and logic, same as AI,” he concluded.
Meanwhile, a sustainability expert raised an issue of data privacy in AI developments. Renata Hartle replied that real estate has always been about the data. Since the industry is accustomed to keeping their privacy guarded, this will likely not change also in the AI’s case. A moment later a recruitment professional asked about the potential for AI to predict the perfect macroeconomic momentums to launch developments, for example. Piotr Trzcinski commented: “It depends on the proliferation of AI power tools for the investment making decisions. I don’t necessarily think it’s going hugely impact the economic cycles as such, but it may increase volatility on the market, if too many parties are becoming reliant on AI power decision making tools.”
Thank you to the ULI Poland Places + Spaces series Partners!
Don’t have an account? Sign up for a ULI guest account.