EP. 338 - Using AI & ChatGPT for Property Advice – The Real Strengths & Weaknesses

Dave opens today’s episode with a deep dive into a trend reshaping how Australians handle money, research, and property decisions; the explosive rise of artificial intelligence. With millions now using AI for budgeting, market summaries, and even suburb selection, the Trio unpack where these tools shine, and also where they can seriously mislead.

Mike shares new research from MCG Quantity Surveyors, exploring whether ChatGPT can reliably recommend investment-grade suburbs across five capital cities. The team put the challenge to the test. The task was a simple everyday prompt, and a highly structured, data-fed workflow using growth rates, listings, vacancy rates, Census variables, and infrastructure notes. Surprisingly, more than half of the model’s “data-driven” answers failed basic fact-checking. It featured the wrong days on market, incorrect price movements, mismatched listing counts and, in some cases, the model selecting suburbs that weren’t even in the dataset provided.

Cate highlights the core issue. AI writes excellent narrative, but it is not a disciplined property analyst. It can summarise infrastructure projects beautifully and generate clear suburb overviews, but it cannot reliably prioritise the drivers of capital growth. It frequently overweights yield, ignores supply-side risk, and treats structured data as just “more text”, leading to confident but brittle conclusions.

The Trio explore why this happens; large language models predict text patterns, rather than test logic or perform weighted analysis. They also considered why this makes AI particularly risky for property decision-making, where nuance, hyper-local variation, and asset selection matter enormously.

They also discussed the growing trend of clients arriving with AI-generated question lists. While this lifts the quality of early conversations, the underlying assumptions often require unpacking.

To wrap up, the Trio outline how investors can use AI safely; as a speed tool for summaries, frameworks, and idea generation…. but never as a substitute for independent verification, professional judgement, or suburb-level due diligence.

Today’s ep is a fascinating look at the future intersection of AI and property.

And our gold nuggets!…..

Cate Bakos’s gold nugget: Cate applies caution to those who don’t question where the data source is actually coming from. AI scoops up all kinds of threads, sales spin and incorrect claims. It’s critical to apply critical thinking.

Mike Mortlock’s gold nugget: Mike tells a little story….. a newspaper story in a paper for Pakistan featured a ChatGPT prompt at the end of the article….ouch. Verifying sources and watching out for those M-dashes is important!

Dave Johnston’s gold nugget: Dave shares a tip he got from a Harvard professor; “Treat AI as just another smart person in the room”

Follow Cate Bakos - Cate Bakos Property
Facebook:   / catebakosproperty  
Instagram:   / cate_bakos_property  
Linkedin:   / catherinebakos  
Website: https://www.catebakos.com.au/

Follow David Johnston - Property Planning Australia
Facebook:   / propertyplanning  
Instagram:   / propertyplanningaustralia  
Linkedin:   / property-planning-australia  
Website: https://propertyplanning.com.au/

Follow Mike Mortlock - MCG Quantity Surveyors:
Facebook:   / mcgquantitysurveyors  
Instagram:   / mcgquantitysurveyors  
Linkedin:   / mcg-quantity-surveyors  
Website: https://www.mcgqs.com.au/