Picture the scene: the founders of an artificial intelligence (AI) proptech tool are pitching a real estate company about helping manage their hospitality portfolio. They draw up supporting data for streamlining operations, cost savings, solutions to staff shortages. The landlord pauses. “That all looks very interesting. But will the AI tell my tenants to kill themselves?”
This sci-fi-inspired scenario actually happened in a European head office in recent weeks, as the messages around AI continue to capture the headlines in ways that were previously unimaginable. ChatGPT has thrust the concept of using AI assistants into the spotlight and it seems like there is no going back. While in some ways, this publicity is helping the wider AI industry to market itself, there are also plenty of negative news stories.
Conspicuously, Tesla and Twitter CEO Elon Musk called for all AI companies to “immediately pause” training AI systems in a letter published by think tank Future of Life Institute on March 22. He tweeted “the singularity is near” on April 15, referring to a hypothetical future scenario where artificial intelligence outwits human intelligence, and technological expansion becomes uncontrollable and irreversible.
Real estate application
How far is this hype, and how should businesses approach the issue? Tom Shrive, CEO and founder of askporter, an AI engagement and task management platform for real estate, concedes that “there are some existential questions that need to be tackled”. However, he points out that askporter is “all about narrowing down issues to solve specific problems that real estate companies are facing”. He adds: “Our platform is designed to ringfence your data. Although we are training a system that can have smart interactions and smart conversations, it’s limited to handling a single asset at a time.”
The platform’s name reflects its origins in residential applications, but its service-led functions have translated into askporter becoming a tool for hospitality assets and many other property types. One key client, a holiday camp operator in Germany, is using askporter to coordinate much of its massive cleaning and housekeeping operations. “The company is using the platform to manage all their cleaner check-ins for individual units,” explains Ben Yexley, the firm’s head of business development. “If you take a holiday park with thousands of chalets, toilet blocks and public spaces, we can tag each ‘asset’ with a QR code; a cleaner then checks in when they have taken care of that space, and it also allows us to create a web assistant for each unit. That way we slowly build the digital twin of an entire portfolio for the client. With other facility management clients, that has become a way too of managing cleaning supplies, processing concierge-style requests, or reporting maintenance issues.”
Yexley underlines that AI is not only very good at handing one-off, menial tasks, but also quite complex repetitive duties. With daily cleaning and maintenance playing such a huge role in any hospitality operation, the quotidian applications are obvious. But does that mean that it has fewer applications in a luxury or bespoke setting, where human interaction is prized? Not necessarily, Yexley says. “Our whole message is not about replacing people with AI automation, but seeing it we can remove the mundane tasks, room bookings, transactions, so people have more time to spend on the valuable jobs. If I am staying in a five star hotel and something is missing from the bathroom, I don’t need to call the front desk, that can be instantly ticketed and solved without a conversation. It’s all about providing efficiencies so that humans can actually add more value.”
Many companies are going in softly with AI tools to solve one or two simple tasks, and that’s just what the firm recommends. Says Yexley: “The way we have built the product is that we start by curating 5-15 per cent of your mundane tasks, and focus on the top five requests that customer service teams get on a daily basis. Then we can build from there.”
AI is also a key interface between non-technical end users and technical specialists. “For example, it can translate a reported technical problem into language that the engineer needs to repair it. It can also help the tenant self serve, so if for example they report that a certain light is flashing on an appliance, the AI, which has read the instruction manual, can directly talk you through fixing the problem,” he says.
Yexley adds: “As AI improves further and the buzz broadens, we want to become the voice and the expert for using AI to make customer service and customer operations better. It will get to a point of effectively managing itself and engage with the data by itself.”
Running into problems
Systems such as ChatGPT are already being used on a daily basis by hospitality and other real estate professionals to write and update property descriptions, craft social media posts and even draft legal documents. Where businesses have run into problems, they have often been lax about company data and proprietary information, or got too “clever” with writing prompts that disable chat safety nets.
But privacy issues are arguably baked into more open tools which learn as they go from information provided by users. Bing AI Chat, Microsoft’s answer to ChatGPT which is currently in limited preview, will attempt to tackle such matters in closed intranet loops, but has warned users to be careful what data they share. Its functionality still has some way to go, in the wake of complaints about toxic and creepy language, reports of assisting hackers, and even threatening to steal nuclear codes.
As concerns rumble on about the impending “singularity” and data leaks, Yexley compares discussions to the advent of the internet. The first Apple computer inspired similar doomsday scenarios of “too much” power in the hands of machines, and a whole slew of menacing 80s movies like WarGames, where a military central computer nearly starts World War III.
Reasons Yexley: “If you don’t try AI for some aspects of company business today, you risk getting left behind. There’s no need to go back to massive call centres with people picking up the phone to solve everyday issues.”
He adds: “We see that we are solving a lot of important problems with bespoke, ringfenced systems and the benefits are clear, although we do welcome ongoing debate about the ethics of it all.”