Mark your calendars for the ultimate real estate experiences with Inman’s upcoming events! Dive into the future at Connect Miami, immerse in luxury at Luxury Connect, and converge with industry leaders at Inman Connect Las Vegas. Discover more and join the industry’s best at inman.com/events.
Since ChatGPT first burst on the scene in late 2022, there has been an onslaught of columns offering advice on how to put ChatGPT to work in your business. While this advice was great for 2023, in 2024, it fails to tap into the power the generative AI provides.
In his talk at ICNY, MIT professor Dr. R. David Edelman, who served as tech adviser to three presidents, explained how we can best navigate what he called the “perilous pivot” that is now occurring where AI has moved from big data and machine learning (its “infancy,”) to the generative AI of today.
What exactly is AI?
When I asked ChatGPT to provide a clear definition of AI for Inman readers, here’s how it responded:
“Artificial Intelligence, or AI, refers to computer systems designed to perform tasks that typically require human intelligence, such as recognizing speech, making decisions, and understanding natural language. It encompasses a range of technologies that enable machines to learn from experience, adjust to new inputs, and perform human-like tasks, streamlining and enhancing various aspects of work, including in the real estate sector where it can analyze market trends, predict property values, and personalize customer interactions.”
Edelman had a slightly different description of what AI does:
“AI classifies huge volumes of information, finds subtle patterns and correlations, learns from behaviors, and ultimately makes predictions based upon machine learning (applied statistics),” Edelman said.
Edelman said the critical thing to understand about AI is that these systems are prediction machines, whether the prediction is what word is going to come next or the prediction of what a certain configuration of pixels means.
In fact, almost everything we’re seeing in the press is about big data and machine learning.
“Most enterprises have just crossed this Rubicon of using big data and machine learning to try to predict a little bit about what they’re doing every day,” Edelman said.
It’s important to understand that “Vanilla AI” has resulted in tremendous breakthroughs in a wide variety of areas, including education, reducing CO2 by using machine learning to predict better bus routes in Boston and preventing deaths from sepsis, where there is a 40 percent mortality rate.
Edelman said that ChatGPT is experiencing what he calls a “perilous pivot” where generative AI is transitioning from its infancy just over a year ago to “Armchair AI” that you can do at home,” Edelman said.
“What I mean by generative AI are these large language models like ChatGPT or Claude and image generators like Dall ·E 2 or Midjourney,” Edelman said. (For a glimpse of astonishing photo-realistic images, follow the Midjourney link.)
These new tools are hyper-advanced and represent the forefront of innovation. What took massive amounts of time, resources, and millions of dollars worth of computers to do in the past is now accessible to you at home for only $10 – $50 per month.
Working with generative AI? Your job description is ‘BS management’
Given the powerful capabilities of generative AI, the percentage of enterprises using generative AI in customer-facing areas is shockingly tiny. Edelman attributes part of the issue to anxiety about the limitations of generative AI. The reason?
“The answer is there is an anxiety, probably a reasonable anxiety, that has crept into the space on the basis of the appropriate recognition of the system’s limitations,” Edelman said.
“The best way to think about generative AI for your enterprise is that it’s like an Ivy League intern who is extremely articulate, supremely confident, and completely full of it ten percent of the time.”
In terms of using generative AI, “Your job now in the world of managing generative AI is BS management, specifically managing to understand what that ten percent is and when that’s happening,” Edelman said.
“Sometimes they [generative AI systems] get it wrong, and the reason they get it wrong is because of inherent limitations in the technology due to how the technology works.”
Challenges ahead
The real-life problems we are already facing in real estate and other areas will explode over the coming years as the use of generative AI continues to spread.
One of the most concerning examples is the so-called deep fakes, where AI can very realistically mimic voices. Scammers are now using these fake voices to verify wire transfers.
“It’s happening all over, and you’ve got to be aware of it,” Edelman said. “The answer is not lines of code, but smart organizational planning, creatively thinking like the bad guys, and recognizing that, especially in industries like yours that are so human, those relationships. Those relationships are as fragile as the technologies” hands that you put it in,” Edelman said.
How do we move forward?
At the beginning of his session, Edelman shared the following quote from science fiction writer Arthur C. Clark:
“Every sufficiently advanced technology is indistinguishable from magic.”
Edelman said that when something seems like magic and people don’t understand it or fear that it is beyond their comprehension, they may become scared or try to cast it out.
“These are systems that we designed ourselves with limitations that we’re aware of that we can actually work through and work around them,” Edelman explained. “They have flaws that we can manage. In other words, AI isn’t magic—it’s math and code.”
Difficult choices that must be made now
Edelman is highly concerned about the choices we need to make right now about this technology and how it interacts with our lives and our livelihood.
We’re in an environment where someone can steal your face and your voice and use your identity without your permission. As we go into another election, Edelman said that one survey showed that only somewhere between two and seven percent of companies are working on political stability or national safety concerns.
“So now at the enterprise level, and at the human level, we must decide what AI should and shouldn’t do, “Edelman said.
The next big step: how to team up with these machines
Edelman’s final piece of advice was to “Team up with AI, be smart about it, and don’t be afraid that AI is going to take your job,” but he then warned, “But someone who knows how to use AI just might.”
For fun, I went to ChatGPT and plugged in this article and asked for its commentary on Edelman’s session. Here’s its take on what Edelman had to say:
“As for commentary on Edelman’s advice, it appears he provides a thorough and realistic assessment of AI’s current and potential impact on various sectors, including real estate. His focus on the importance of understanding AI’s predictive nature and the distinction between generative and traditional data-driven AI is crucial for real estate professionals.
He rightly points out that AI should be viewed as a tool rather than a replacement for human expertise, emphasizing the importance of managing AI’s limitations while leveraging its strengths.”
Edelman provided a road map in terms of what generative AI can do, but how do you prompt it to obtain the best results for your real estate business when the old strategies for 2023 fail to tap into the full power of generative AI
Bernice Ross, president and CEO of BrokerageUP and RealEstateC
