AI Nearly Kills $50M Deal, Proving Agent Value

AI Nearly Kills $50M Deal, Proving Agent Value

A $50 million real estate deal was nearly derailed when both the buyer and seller turned to ChatGPT for advice. The seller asked if they should sell at the offered price, and the AI suggested no. The buyer asked if they were overpaying, and ChatGPT agreed that they were.

This situation highlights a growing trend where people are using artificial intelligence tools like ChatGPT for major life and financial decisions. At the same time, the real estate industry itself is adopting AI, creating apps that help buyers search for homes and neighborhoods through simple conversations.

Agent Recovers Deal After AI Missteps

Luxury broker Ryan Serhant shared the dramatic story, explaining how he was able to salvage the $50 million deal. It took him about a day and a half to fix the situation. Serhant noted that the objections raised were based on information taken as truth, rather than as a validation of how questions were asked.

“Agent beat AI this time,” Serhant quipped, emphasizing his role in overcoming the AI-driven obstacles.

Serhant explained that he resolved the issue by teaching both the buyer and seller a crucial point about AI. He showed them that large language models (LLMs) like ChatGPT frame their assumptions based on how a question is asked. The AI then optimizes for a coherent answer, not necessarily an objectively correct one.

For example, if a buyer asks if a price is too high, ChatGPT tends to lean towards caution and find reasons why it might be overpriced. Conversely, if a seller asks if a price is too low, the AI will likely take an opportunistic stance and highlight potential upside. AI can model the market, but it cannot model the specifics of a deal.

When both sides use the same tool to get opposing answers, it shows why human expertise is more valuable than ever. Agents can interpret information and understand context in ways that AI currently cannot. This is what they do all day long.

AI’s Reach and a Happy Ending

Serhant’s video about the incident went viral, reaching about two million views. It was shared widely, including with the buyer and seller involved in the deal. Seeing that other buyers were circling and that the deal might proceed without them made both parties nervous.

This unexpected reaction from the humans, something AI cannot anticipate, ultimately helped. The buyer signed the deal that morning, and the seller signed just 17 minutes before the conversation. This led to a happy ending for the $50 million penthouse apartment sale in New York.

Beneficial Uses of AI in Real Estate

While the near-failure of the deal highlights potential AI pitfalls, Serhant also pointed out beneficial uses of AI for homebuyers and sellers. He sees AI as an additional tool in one’s toolbox to help frame and model decisions. However, it should not be the only tool making the decision.

At Serhant, the company has developed an AI called Simple. It is designed specifically for real estate and has raised significant funding. Simple acts as an agentic artificial intelligence, providing a deal analyzer.

With Simple, users input all available information from the AI and the internet. Crucially, the human user controls the model’s inputs. They can add proprietary comparable sales data, off-market listings, and internal company data that the public doesn’t have access to. This allows for a standardized way for clients to evaluate pricing.

Serhant believes that by using AI correctly, companies can win deals instead of losing them to AI. It is an amazingly helpful tool when used properly.

Market Impact

The incident involving the $50 million deal serves as a stark reminder of the limitations of artificial intelligence in complex financial transactions. While AI can process vast amounts of data and provide quick answers, it lacks the nuanced understanding, contextual awareness, and human judgment essential for successful negotiation and deal-making.

This situation underscores the continued importance of skilled professionals, like real estate agents, who can interpret AI-generated information, manage client emotions, and navigate the intricacies of a deal. The ability to understand how questions are framed and how AI responds is a critical skill that human agents possess.

For investors and consumers, the key takeaway is to view AI as a powerful assistant, not a sole decision-maker. Tools like ChatGPT can offer valuable insights and data points, but they should be used to supplement, not replace, human expertise and critical thinking, especially when significant financial stakes are involved.

The real estate industry’s adoption of AI, as seen with Serhant’s Simple platform, suggests a future where AI and human agents work together. This collaboration could lead to more efficient processes and better-informed decisions, provided the human element remains central to guiding the AI’s output and managing the client relationship.

The long-term implication is that AI will likely become more integrated into financial decision-making processes. However, the value of human advisors who can provide context, manage risk, and offer personalized strategies will remain paramount. The ability to ask the right questions and critically evaluate AI-generated answers will become an increasingly important skill for everyone.

Source: ‘Agent BEAT AI this time': Real estate agent explains how a $50M deal was almost derailed (YouTube)