Rebuilding broken trust in the age of AI

Transcript:

Allie Barefoot: Welcome back to Data in Context. I’m Allie Barefoot with Cotality. Trust is easy to lose, and almost twice as hard to gain back. In the last year, the housing market’s relationship with AI has moved from a honeymoon phase to a reality check. In our most recent survey here at Cotality, our data shows that while people acknowledge that errors happen, not all errors are held to the same standard. While we found that most people forgive human error, there’s little forgiveness for machines. 70% of those surveyed by Cotality say a single AI mistake would break their trust in a platform. The future of real estate isn’t just AI; it’s AI authenticated by humans. By keeping humans in the loop, we transform algorithmic anxiety into verifiable confidence. So today we’re talking with Cotality's Head of Research and Development, Anand Srinivasan, to talk about moving away from unexplained logic and toward verifiable insights that keep the human in the loop. Let’s go ahead and jump into today’s questions with Anand. Anand, welcome back to Data in Context. So happy to have you here again.

Anand Srinivasan: It’s wonderful to be here again and talking about a survey that we did this time last year. It’s amazing how quickly time flies.

Allie Barefoot: I know! We were just talking about how quickly this year has already flown by, and the fact that we’re already talking about another topic again evolving around the topic of trust. So I want to jump in here, like you said, to that survey question. Anand, Cotality’s survey showed that perspective homebuyers' trust in AI platforms dropped by nearly half—from 30% to 16% in just one year. What happened in those 12 months to cause such a massive trust cliff in a market that was previously optimistic?

Anand Srinivasan: Look, Allie, AI is new, and I think that with any new technology, any new process, any new system, trust builds over time. Um, and it doesn’t build in a straight line either, so we’re going to see some ebbs and flows. We think AI is here to stay; we think that the usage of AI in the property ecosystem is likely to increase, but it’s not going to be a straight line from a trust perspective. And we’re seeing a little bit of that. So, in a lot of ways, this isn’t completely, um, uh, surprising, if you may. Um, but we’re going to see—we’re already starting to see some pockets of that. If you peel the the layers of the onion of the data, uh, a little bit, we’re starting to see that. But look, we’re going to see some ebbs and flows, but I think AI—trust in AI—will come over time as it proves its worth and it proves its value.

Allie Barefoot: Oh yeah, just like anything that’s new, you know? You gotta practice with it; you gotta take some practice swings before you get in the batter’s box there.

Anand Srinivasan: Absolutely.

Allie Barefoot: And we’re seeing a little bit of a double standard here where users uh allow humans to have some error, little bit of cushion when when they make a mistake, but they demand a digital platform like AI to be perfect. So why are we so much more likely to abandon a platform over an AI error as opposed to a human one?

Anand Srinivasan: Look, we want—we have seen the prowess of technology, we have seen the prowess of computing, and we are, um— we can understand what it is capable of. So, our— we have zero empathy when it comes to AI. We have and will continue to expect perfect results, or near-perfect results. On the other hand, we recognize that humans are fallible, and we have seen them make mistakes, but we trust them implicitly, particularly if we have worked with them before. And as a result, there’s a little bit more give in the system from consumers towards their um human property ecosystem counterparts. On the other hand, like with anything else, the there is no mercy, if you may, towards uh technology and AI. So again, not surprising in the grand scheme of things, uh, but this will build up over time and it will, um, propagate across multiple prop- parts of the property ecosystem, and um, and we will—we will get used to incredible efficiency, um, incredible time savings, potentially cost savings, and uh, we will also recognize that the human in the loop is a significantly important part of that ecosystem.

Allie Barefoot: 100%, you know? There is a different level of trust throughout generations when it comes to technology in general, not just AI. And the data shows, Cotality's data shows, that baby boomers are 14 points less forgiving of AI errors than Gen Z—who obviously are a little bit more digitally native, they grew up with it. They also maintain an 8-point caution gap, so they're not fully invested in trusting this platform just yet either. Is this because the trust isn't about tech literacy, or is it more about the sheer weight of a financial transaction when it comes to buying a home?

Anand Srinivasan: Uh, I think—I think you’ve nailed it, which is that I think baby boomers have been through multiple transactions, um, over the course of their lives, and they have done—they’ve seen some uh very good transactions, and arguably they’ve had some mishaps along the process that has solidified their their view on this process. As a result, they are less forgiving um on mistakes, and at the same time they recognize, or perhaps because of the fact that it is such an important decision in their lives, that they place an enormous weight on it and are less forgiving. On the other hand, if you look at the younger generations, um, a, they haven’t been through the process as many times as the boomers, number one. So there’s a a little less experience when it comes to that. Second, um, they’re familiar with tech, but they haven’t had tech plus a huge financial decision um that could last multiple decades combined together. So they haven’t had that experience just yet, and once they—once they do, they will build up that knowledge repository. But what’s different between them and the and the boomers is that AI is getting integrated into their processes across the board—whether it’s the property ecosystem, whether it’s other parts of their lives, whether it’s work, whether it’s personal. AI usage is here to stay, and we’ve seen that from the survey data. So when it’s so tightly integrated, they will um likely accept it as part and parcel of the property ecosystem, and their trust in it will build up as they have successive um positive experiences.

Allie Barefoot: And let's talk about that "one strike, you’re out" type of narrative that some people have with AI. When 70% of people say that a single AI error breaks their trust, how does Cotality integrate transparency to prevent one mistake from destroying a long-term relationship?

Anand Srinivasan: Look, I mean, this is um uh part and parcel of our of our DNA and how we do things. We take the data accuracy and completeness and thoroughness, uh, both from a system standpoint as well as a human-in-the-loop standpoint, very, very seriously. We recognize how important this data is, and more importantly what this data means, um, in financial context and uh how big of a role it plays in people's lives. We’re keenly aware of that connection. As a result of that, we put an extra oomph and then some into validating the data behind the property ecosystem, right? So, I think that it plays really well to how we approach the property ecosystem with respect to AI from a process perspective. So when we use AI, we're using it incredibly um deliberately, um, in a um in a very thoughtful way, and um, you know, trust but verify, um, we do that time and time and time again. Um, in fact, we’ve been using um some sort of machine learning for a decade-plus now as part of our processes, and we have been very transparent about it. We know exactly where it’s being used, we articulate where it’s being used, how it’s being used. Remember, we're a pretty regulated industry, so it's important that our customers and stakeholders recognize how and when and where we’re using uh AI in all of its forms. So, I think that it’s important to articulate the the accuracy and the completeness and the time uh efficiency and the cost savings it brings to the table, but we also recognize that we have uh responsibility in being transparent to all the different stakeholders as to how and where and when uh AI in all of its forms is being used. So we do that very, very thoroughly.

Allie Barefoot: And because we see that this trust is slightly falling off a cliff when it comes to AI, you’ve mentioned that Cotality’s principles are always keeping a human in the loop. How important is that with creating a bond to keep that human in the loop to make sure that if there is that AI error, there’s a human there to back up and explain that further? How important is that human-in-the-loop scenario?

Anand Srinivasan: It’s critical, right? We’ve seen that from the data across generations. This is—look, this is the most important decision um many of the, particularly the younger folk, will make in their lives as far as they have been through that stage now, right? So as of that point, the 30-year-old, the 40-year-old, this is the single biggest important uh financial decision they will make, and it’s been driven by life events um, you know, family, moving to a different neighborhood, schools, etc., etc. Super critical to have somebody who understands uh the technology and how it's being used in the different uh pieces of the puzzle there for sure, but also articulating at a human level to uh the consumers of this that, look, this is how the data has been used, this is why we use that, this is what the data means for you, and this is how it shapes um and is part of their decision or choices that they have to make um in terms of the property they’re going to buy, the location they’re trying to move to, the different financial uh pieces of the puzzle they have to figure out. So, to answer your question in a terribly long-winded way, um, but super critical uh to have a human in the loop.

Allie Barefoot: No, absolutely. And my last question here for you, Anand, is: how do we move past the culture where technology produces results but doesn’t explain how? What does a "trust but verify" future actually look like for the average homebuyer?

Anand Srinivasan: Look, transparency is part of the process. Um, it ought to be part of the the culture of the company and the ecosystem of companies that you’re dealing with that this notion of um uh data integrity, process integrity, technology integrity, it is core to the thecompany that you're dealing with. And at Cotality, I can state, um, without any qualms, that it’s unequivocally core to our DNA. We take um an incredible amount of effort and um a sense of pride in being able to say our coverage is very high, our quality is very high, and um, we also provide that transparency. So it’s—it’s super important both from a data perspective as well as from a process perspective to be able to share that with the consumer—regardless of where they are, whether it’s their first house or their tenth house, whether it’s a minor financial decision or a major financial one, whether they’re up-sizing or down-sizing their home. It doesn’t matter. We have to be thorough and consistent in articulating um that "trust but verify" scheme that you mentioned, Allie.

Allie Barefoot: Anand, this has been truly eye-opening to see the data in how we’re losing a little bit of the trust in AI just as a society, but knowing that when you have transparency, it can definitely make it a little bit more digestible. So I’m very excited to see what conversation we will have a year from now, as it’s been one year since we’ve talked about AI originally. Anand, thank you again for joining me on Data in Context.

Anand Srinivasan: Great to be with you, Allie. Thank you for your time.

Allie Barefoot: Thank you again to Anand Srinivasan for joining me here again on Data in Context, and thank you guys so much for listening. If you haven’t already, please subscribe to Cotality's YouTube page. And if you want to find out more information, as always, head on over to cotality.com.