"We all need to sleep," says Zsolt Kohalmi, Pictet Alternative Advisors' global head of real estate and deputy CEO, as he discusses the inescapable impact of artificial intelligence on the future of the commercial real estate sector.
The truth is that AI is already influencing real estate decisions, with Kohalmi and his colleague Partha Sarathy, head of asset management, highlighting how data centres have captured the attention of global investors over the past 14 months, alongside political headlines.
Speaking at a roundtable breakfast at the company's West End offices on 5 Stratton Street in London this week, Kohalmi labelled AI as an "interesting" scenario for the real estate sector, presenting both challenges and opportunities.
His comments followed an eventful week for the industry after the heads of three of the world's largest real estate consultancies found themselves addressing concerns about the potential effects of AI on their businesses, despite posting encouraging financial results this month, as reported by CoStar US.
Addressing the elephant in the room, he insisted that real estate should face less interference from increasing AI adoption compared with other industries.
"The last few weeks, you have seen what it has done in the stock market, right? Everybody seems to change their mind by a portion of 20% as to what a company is worth depending on their vision of how AI will impact it. I think it certainly feels like a relatively historic time for all of us to try and understand.
"Within that context, I think one interesting point is that real estate is one of the few major sectors around that one can argue will face limited disruption from AI; it doesn't apply universally, but when you look through the sectors, it really is one of the very few areas that can claim that."
Despite shared uncertainty around how AI's development might affect businesses and communities over the next few years, Kohalmi highlighted that some real estate was more likely than others to be safe from major disruption, relating back to his previous comments sleep.
"At least today, we don't see any solution to that problem," he said. "So I think that the residential side of all of this, and anything beds-related is very safe.
"If you think that AI will mean that perhaps people will work less, than we better figure out more hospitality because people are suddenly going to have more free time. So, I think anything that is related to beds is going to stay."
Speaking with CoStar News later in the session, Kohalmi said Europe's status as a desirable tourism destination should give the sector, and therefore investors, a sense of certainty, adding that the continent should "up its game" in the area in reference to its undersupply of hotels.
He also flagged student accommodation and senior living as other possible winners from greater AI adoption.
Need for places
During the main session, Kohalmi said logistics was another sector that was less likely to feel a disruptive effect from AI advancement.
"There is some talk of more efficient logistics, yes, but again I think, particularly last-mile logistics near cities given this decreasing supply as they grow, these areas feel very good in a world of AI where it is new territory for companies to see companies valuations affected by up to 20% in one day."
But he warned that some areas of real estate would be less fortunate, providing caveats to his earlier comments that the industry, as a whole, will be one of those least affected due to the 'undisruptable' quality of people needing places to live, work and spend time.
Offices were one of these, with AI already streamlining workforces at major companies and more expected, threatening a knock-on effect on floorplates. "I think [AI] will be another road bump," the Pictet deputy CEO said.
"Today, in every conversation I have with institutions, people are becoming more careful around their actual HR needs going forward given the impact of AI and so, to me, it is very hard to imagine that is not going to have an impact on offices.
"What I would say is that it is going to be very different because what it will definitely mean is that locations like the one we are sitting in today will be fine, no matter what happens... places like Mayfair or central Paris will be fine but, do we need the floorplate office space that we needed previously? That is very unlikely. The question for everybody to answer is where is that cut off between the two?"
Secondary offices opportunity
Focusing on that point, Kohalmi said that secondary offices would need to be reimagined to improve their investment appeal when the demand for them reduces.
This, he said, would leave the door for developers and investors to explore alternative uses for these sites, with residential being the obvious candidate. "That trend is going to come for sure," Kohalmi said.
"I think anybody that buys an office building today in a slightly secondary location with a seven-year lease is taking an incredible gambit because, you buy an office building like a fixed income instrument, because you believe it is an eternal cashflow.
"But seven years from now I don't think any of us have the power to say how the demand for offices will be in secondary locations or what type of floorplates companies will need, I think that is a tough call.
"It would be more suitable for a value-add investor who has an alternative use in mind. But most core buyers will not have an alternative use and the ability to do that in mind when they are buying an office building today."
"Seven years from now I don't think any of us have the power to say how the demand for offices will be in secondary locations."
He added: "So I think office buildings become much more restricted in location if you are looking for certainty just because of how long a lease is and how much [of an] impact AI may be able to have in these coming years."
Another area of real estate Kohalmi also raised uncertainty about was data centres. Despite arguing that "data was exploding", with major investors pivoting to deploy capital into data centres over the past year, he said questioned future demand for such structures.
"The data centre theme will very much depend on how many Large Language Models do we really need, which is a many trillion dollar question."
He added that another potential headache for the data centre sector was who they would be sold to in the future, with the buyer pool uncertain today.
"The biggest question for anybody doing one of these large data centre deals is where is the liquidity, who will this be an object of desire to, who will be the large buyers and I think, if you are doing anything in four or five or six years time, that is not a question we have a good answer to."