Can AI design a dream home? This is what ChatGPT says Australians want, based on Domain data

Ever wondered what the collective dream home of Australians might look like?

Is it a Queenslander house with picture-perfect verandah plucked straight out of Bluey? Is it a red-brick art deco home in one of Melbourne’s beachside enclaves? Or perhaps, a harbourside studio apartment in Sydney with postcard views that look like a Ken Done painting?

It turns out it’s none of them.

We asked ChatGPT to design Domain readers’ dream house using the most searched keywords across our property portal.

A dream home is many things to many people, so in the spirit of being as fair as can be, we used Domain’s top 20 keywords across all capital cities, as aggregated by Domain’s research team.

We then punched these keywords into ChatGPT like vegetables in an artificial minestrone and waited for an image of our dream home to simmer from behind our screen.

For budding Ian Thorpe hopefuls, the number one searched term across the nation was “pool”.

A north-facing aspect, waterfront views and penthouse vibes were also favoured keywords by Domain users. Must-have amenities included a garage, courtyard and study.

The resulting “dream home” looks like a modern two-storey, double-fronted waterside property on the surface.

Look a little closer, and a blooper reel of awkward truths starts to appear. Just ask Simone Azzi, associate director of Belle Property Annandale.

For one, is this house even sellable?

“In Australia, probably not,” Azzi says. “That pool does not look like it’s compliant at all.”

Even if the pool did meet safety regulations, its placement next to the driveway in the front garden would be a hard sell – and it would be an even harder journey for those steering their car up the narrow driveway, where an unforgettable, costly and litigious car wash awaits.

“The location of the grass and the pool is just wrong, wrong, wrong,” Azzi says.

“Nobody wants to walk out of their front door and jump in the pool.”

While appraising the home, Azzi looks for an optimistic spin that might pique buyers’ interest.

“On a positive note, it’s double-fronted and has a garage, which is always attractive,” she says.

“There are plenty of windows to maximise light, and from a design perspective, it’s incorporating curves, which are very on trend at the moment.”

The aspect of the home, though,is a missed opportunity, Azzi adds.

“The house should be facing the water to maximise the view from some of those living spaces,” she says.

While the house does possess some surface-level eye candy, overall, Azzi doesn’t consider it to be practical.

“It looks lovely, but it is not a practical, functional home; it’s got lots wrong with it, so thanks ChatGPT, but no thanks.”

And no, not even the “medieval-but-make-it-modern” moat that surrounds the property can convince her otherwise.

For a little more context, we also asked Chat GPT to design the Domain audience’s dream house in Sydney and Melbourne, using keyword search data specific to each capital city.

The results might not convince buyers to stock up on moving boxes and masking tape, but they do reflect the key property differences and preferences between each city, says Domain chief of research and economics Dr Nicola Powell.

“If you put those three homes together, it really showcases the different needs and wants in our different capital cities, and what buyers prioritise in their homes depending on the city they live in,” she says.

The Sydney home is a three-storey, harbourside property that would fetch a price range that is miles ahead of the city’s current house median of $1.7 million.

Local agent Simone Azzi estimates a waterside house with these proportions would fetch upwards of $8 million.

“If it’s got water, it’s that big and it looks like it’s a decent plot of land, I would say in metro Sydney it’s got to be … $8 million to $10 million-plus,” she says.

But securing a sale with another pool in the front yard?

Most agents are going to have to work extra hard for their commission with this unusual quirk.

“I don’t know what’s going on in ChatGPT’s world with these pools in the front yard. Who does that?” Azzi asks.

On the plus side, the home’s arched window, curved silhouette and harbourside views would be favoured by buyers, she adds.

In Melbourne, “art deco” ranked as one of the most-searched words by house hunters, and the resulting AI-generated home has enthusiastically embraced this prompt.

Powell is not surprised the “dream home” leans closer to a period home than a new build in the city.

“Melbourne has a really strong following of heritage and art deco homes,” she says. “It comes at a premium, but it’s the creme de la creme of Melbourne real estate.”

However, it’s the inaccessible Juliet balcony that raises the most questions for Powell.

“It really demonstrates the blooper that AI can create,” she says. “You would not have a Juliet balcony like that hanging over a pool, and there’s no accessibility.”

When asked which of these three homes she’d live in, Powell is quick to pick the Sydney property. But she’s happy to leave the house behind in its pixel format and instead has her eyes on the waterfront location.

“A dwelling like that on the harbourfront of Sydney is becoming rarer and rarer,” she says.