TL;DR
Get home appliances delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
Apartment Therapy writer Cori Sears tested ChatGPT, Google Gemini and Claude for ideas to restyle her living-room shelves. She selected ChatGPT’s image suggestions to inform a refresh using decor she already owned. The report describes a personal experiment, not a controlled comparison.
Apartment Therapy writer Cori Sears tested ChatGPT, Google Gemini and Claude for ideas to redesign her living-room shelves. She used suggestions from ChatGPT to guide a holiday-season refresh with decor already in her home. Her account describes how the platforms responded to the same request and how she adapted one set of suggestions to her room.
Sears uploaded a photo of her built-in shelves and gave all three platforms the same prompt: she wanted the shelves to look more aesthetic, with a minimal, timeless seasonal touch, and asked for inspiration images. She later said the wording may have left too much room for interpretation: she wanted the holiday decor to be minimal, but that preference did not come through clearly in every result.
Sears selected ChatGPT’s results because they included several visual variations she could adapt. She also said its renderings could distort size and proportion and look heavily filtered, making them difficult to reproduce exactly. She said Gemini’s mockups did not suit her, including after she tried revising the prompt. Claude, on the free version she used, did not generate a custom image; it supplied web images and written guidance instead.
To put the ideas into practice, Sears cleared and wiped down the shelves, then returned some items that appeared in ChatGPT’s mockups, including a lamp, framed picture, potted plant and wicker basket. She drew on her existing holiday decorations and vintage and antique finds, rather than shopping for new pieces. The supplied report excerpt ends as she describes adding seasonal decor, so it does not establish a complete account of the final arrangement or its outcome.
AI Ideas Meet Real Shelves
The test illustrates a distinction between visual inspiration and a practical plan. A generated image can depict a possible arrangement, but may not account for fixed electronics, actual object dimensions or the need to keep everyday items accessible. Sears had a modem, TV remotes and her husband’s PlayStation among the objects on the shelves, but did not include those constraints in her prompt.
The account also describes how the prompt and platform affected the suggestions Sears received. She adapted the generated concepts using decor she already owned. The report does not measure whether using AI was faster or more effective than other design methods.
Products Worth Considering
Blueair Blue Pure 211i Max Air Purifier for Large Rooms
Hisense 75U6SF Pro 75-Inch U6 Pro Mini-LED 4K UHD Smart Fire TV
Philips Hue White and Color Ambiance A19 Smart Bulb Starter Kit, 2-Pack
As an affiliate, we earn on qualifying purchases.
A Shelf Styling Challenge
Sears described the built-in shelves as a long-standing design problem. She had tried collecting inspiration and arranging personal decor, but felt the result remained unfinished. The shelves also had to accommodate practical household equipment, creating a mix of display and storage needs.
With the holiday season approaching, she wanted to refresh the area without buying new decor. Rather than relying only on the AI platform she usually used, ChatGPT, she compared it with Gemini and Claude using the same request and a photo of the room. The report is a personal test, not a formal product review: it does not describe standardized scoring, multiple rooms or a broader user trial.
“I don’t use AI much — but when I have, I’ve usually turned to ChatGPT, so it was my first choice.”
— Cori Sears, Apartment Therapy writer
Limits of the Home Test
The report reflects one person’s experience with the platforms and the versions available to her. It does not establish how the tools perform across different rooms, prompts, subscription plans or later product updates. The supplied excerpt also stops before Sears fully describes the completed shelf arrangement, so the final visual result and her overall satisfaction with it are not confirmed here.
It is also unclear whether the platforms could have produced different suggestions if the prompt had specified which objects could not move, the shelves’ dimensions, or that existing decor should be used. Sears acknowledged that she left some of those constraints out, so the test did not assess how the tools respond to a fully detailed brief.
The Final Arrangement and Follow-Up
The source material describes Sears beginning the redesign and gathering pieces from her home, but does not provide the full account of the finished shelves. It therefore does not establish how closely the final arrangement matched ChatGPT’s images, how it accommodated the modem and PlayStation, or whether the result remained functional.
For readers trying a similar exercise, the reported test points to including fixed equipment, dimensions, budget and available objects in the prompt, then treating generated images as references. Sears’ account does not establish whether this approach would produce more accurate results; those details would address constraints she says were omitted from her initial request.
Key Questions
Which AI platform did Sears prefer for the shelf redesign?
Sears selected ChatGPT because it offered several visual variations she could use as inspiration. She said the images could misrepresent proportions and look heavily altered.
What did Google Gemini produce?
Sears said Gemini’s mockups did not match what she wanted, even after she adjusted the prompt. That is her assessment from this personal test, not a general performance finding.
Could Claude generate a custom shelf image?
On the free version she used, Sears said Claude could not generate a custom image. It returned web images and written styling guidance instead.
Did Sears buy new decor for the redesign?
No. Sears said she intended to use items she already had, including holiday decor and vintage or antique pieces from her collection.
Does the report establish that AI improved the finished shelves?
No. The available report excerpt describes the test and the start of the redesign, but does not provide enough information to confirm the completed result or compare it with a non-AI approach.
Source: apartmenttherapy.com
