REAL AI: Debunking ChatGPT Myths, AI and Data Standards at RESO, Facts, Headlines and AI Quote of the Week

By Kevin Hawkins with Korey Hawkins | Vol. 2 Post 43

REAL AI is a human-created weekly roundup of all things related to artificial intelligence in real estate and emerging AI innovations in other sectors likely to impact our industry. We post a new edition every Friday, and our free newsletter is delivered every Monday.

Debunking the Top 3 ChatGPT Myths

For those born post-internet, a famous comedian named Rodney Dangerfield coined the line “I don’t get no respect. No respect at all” as a catchphrase of his standup act.

Often, AI is maligned just like Rodney. Some of the criticism is spot on. Both other mistruths are either outdated or mischaracterized.

Many AI myths permeate often because of the lightning speed at which it seems AI is progressing – particularly Generative AI – it’s important to take a Snopes approach to some of the falsehoods and misstatements that continue to make it to the board rooms and on stage at conventions.

Myth 1: ChatGPT sucks at math. Misleading. This is one of the most common misconceptions about ChatGPT today. Even at a recent, reputable real estate tech conference, an AI speaker suggested that large language model tech, like ChatGPT, struggles to do “really basic math.”

Because it is a language model, ChatGPT was not explicitly designed for complex mathematical problem-solving. But that doesn’t mean ChatGPT 4o is poor at math.

The reality is that ChatGPT 4o can perform math calculations correctly. It can handle a wide range of real estate-related math tasks that agents face regularly, whether it’s calculating percentages, mortgage payments, or even helping with property tax estimates. If you need a quick estimate of your client’s monthly mortgage payment based on a loan term and interest rate, ChatGPT can assist.

ChatGPT 4o can perform complex math calculations by writing and executing Python code, which is one of the ways it has evolved to address the math-related limitations of earlier versions.

This capability means that for more intricate or multi-step calculations, ChatGPT leverages Python, allowing it to perform tasks accurately, such as helping calculate survey findings and including crosstabulations, mortgage amortization schedules, or property investment projections.

Myth 2: ChatGPT is only good at content creation. False. ChatGPT 4o Gen AI can perform a wide range of non-content creation tasks. You can do dozens of things with ChatGPT 4o beyond getting help writing property descriptions, blogs, and social posts. ChatGPT 4o is excellent at crunching survey data, extracting data from images, organizing unorganized text, helping you calculate expense reports, and much more.

One of the most popular ways to use ChatGPT 4o for non-content purposes is to load the app on your phone and use it instead of Google Search. Open your ChatGPT app on your iPhone or Android device and dictate your search. ChatGPT’s capability to take incomplete pieces of information and find the answer will surprise and delight you. You could type in a half dozen searches and see the same result on Google or do a single prompt with ChatGPT.

Myth #3: You need to be tech-savvy to use ChatGPT. Untrue. Sometimes, the hardest thing to do is to convince non-techie agents to try and use a Gen AI tool like ChatGPT. I don’t blame them. For years, many technology companies have overpromised and underdelivered in providing tech tools for agents that are easy to use. Many are just too complex.

The same remains true for many AI tools that target agents today: they are too complex to use unless you are tech-savvy but give the illusion that anyone can use them. Very few agents will “build” their own GPTs – so why bother to suggest that as many real estate speakers do? It just causes confusion and disappointment in the marketplace.

But ChatGPT is different, especially today’s paid version, ChatGPT 4o. If you can use Google to search for something, you can use ChatGPT. That’s about the only skill level you need to get going.

And that’s the most important part: whatever your tech skill level, minimal to high, diving in and getting used to it is the best and fastest way to learn. Just the time you’ll save searching for quick answers will save you enough time in a month to more than pay for your $20 investment. (-Kevin)

AI and Data Standards at RESO

My first RESO (Real Estate Standards Organization) meeting was almost a decade ago. Marilyn Wilson and I at the WAV Group were hired to help expand the organization beyond the Geek-only crowd that comprised its membership.

The 2015 Conference was at the NAR headquarters in Chicago in a meeting room with pillars that blocked my view of the speaker podium. There was no stage, and I counted three women in the room. RESO had 195 members, and probably one-third were in attendance.

Today, over 1,000 firms are members. The Fall 2024 Conference packed the large ballroom at the Omni Star in Frisco, Texas, located next to the NFL Dallas Cowboys practice facility.

As with nearly every real estate and tech conference today, AI was center stage at RESO.

The one AI session that stood out for me was with Dominik Pogorzelsk, GM of MLS for Restb.ai, the leading computer vision firm in real estate. I know Dominik (full disclosure: Restb.ai is a WAV Group client), but I never saw him on stage before. He was polished and poised, and he had a powerful message that directly connected the importance of what RESO does to making AI the best that it can be.

Dominik concluded his talk with this takeaway: “Clean, standardized data is the key to unlocking the true power of AI in real estate.”

Data is the fuel for AI, and high-quality data can turbo-boost the performance of AI: it’s quality in and out. (-Kevin)

AI Facts

52% of support leaders are worried that they are falling behind in using AI solutions – Intercom
74% of service reps believe AI tools can quickly find and use information about customers – Zendesk
62% of customer support specialists think AI/automation can help them understand customers better – HubSpot
79% of businesses think automation is essential for a good customer experience – Verint
Almost two-thirds of customer service tasks can be automated with an AI-powered system – McKinsey

Source: Fluent Support (-Korey)

AI Headlines

AI Within Reach | 10/21/24 NAR By Kevin Hawkins
How AI is being used today by MLSs and Associations.

Scammers are stealing homes from under their owners’ noses. AI is making it scarily easy | 10/22/24 Business Insider
AI tools are now being used to fake deeds and transfer homeownership away from their rightful owners.

9 ways agents should be using AI to drive real estate sales | 10/22/24 Inman
How to integrate AI into your real estate workflow.

Introducing computer use, a new Claude 3.5 Sonnet, and Claude 3.5 Haiku | 10/22/24 Anthropic
Claude AI can take over your mouse and control your computer!

Microsoft Photos update with super resolution begins rolling out to Windows Insiders | 10/22/24 Microsoft
Microsoft Photos uses AI to take a blurry image and create a clear, high-res version.

BONUS – MUST SEE video: Runway’s Act-One
https://runwayml.com/research/introducing-act-one (-Korey)

AI Quote of the Week