As a cabinet designer, there are numerous hood options to offer from metal to matching cabinetry. However, when a client asks for a simple hood made from stone, it becomes a little trickier. Yes, there are ways to design it in a CAD program, but those are time consuming and take a mid-level of CAD expertise. And in reality, this is being sold/made by someone else, so spending a lot of time on design specifics doesn't make sense (unless you are charging consulting fees to design the hood). In this example, a simple block structure was created and then using AI, it was arched and colorized, extractor insert and pot filler added. Maybe not the final exact specifications, but close enough to convey the design. Does it work to show the design direction? How do you feel about hoods that go to the counter?

While some people try and use AI to completely generate a project, I've found that it takes more time to get the sizing correct in a kitchen where things are measured to the fraction of an inch.

About the design.... this shows a Sub-Zero 48" fridge flush inset into a 50" opening with extended depth. Also something different is rotating the end cabinet 90 degrees for a better interior display (shallow/wide instead of deep/narrow). The tall panel to the right of the cooktop is actually a push-in door that goes into a hidden Chef's kitchen.

For this custom home, I did the layout in the cabinet manufacturer's recommended design software. That way the underlying pricing and part numbers are correct. But that software, only goes so far towards realism. The challenge is when to export the image, and enhance it with AI. In this case, I left out the doors, overhead lighting, decorations, texture and lighting enhancements.

Notes: The AI in this was rendered in an awesome tool - https://mnml.ai/

Because the files are large and not SEO friendly, used free software Irfan to rename and compress files - dropping individual file size from 8mb to 250k at the same pixel size.

Starting from a basic architectural perspective, I used AI to explore how different material directions can completely transform the same kitchen.

All four versions share the exact same layout and viewpoint—only the finishes, tones, and level of refinement change.

Directions explored:

• A. Light / shaker
• B. Warm wood
• C. Dark / high contrast
• D, Mixed /mid contrast

Also included: the original base rendering used to generate the concepts.

AI tool: mnml.ai

💬 Question

Which direction feels most resolved—and why?

• What feels the most livable?
• What would you change or push further?

Testing how subtle vs more expressive AI enhancements affect the perception of a modern home exterior.

Which feels more believable? Which feels more compelling?

Because of security and privacy concerns, we wanted to test whether changing the garage door from transparent glass to something more opaque changes the curb appeal of the home. And to give it a neighborly touch, we had AI add the house number to the sidewalk edge. Lastly, we used two competitive AI tools to meet the challenge.

Do you think the garage door change was needed, or did you prefer the clear glass garage door? Do you feel mnml.ai or ChatGPT did a better job with the changes? Do you notice anything odd when zooming in on the mnml.ai version?

Exploring how AI changes perception through lighting and material refinement in wardrobe design.

Which version feels most livable? What would you change?

In the attached four images there are two of a modern wardrobe from the outside and two of the insides with doors removed. Of each group, one is of a standard CAD perspective drawing and the other is an AI enhanced.

Exploring how AI refinement changes the feel of a transitional kitchen.
Same layout, different levels of detail and material expression.

Which version feels most resolved—and why?

In a community that focuses on kitchen remodeling, the homeowner was asking if they should paint these grey cabinets. But this same scenario could come to a designer in a showroom. So, to showcase upgrade ideas, I did four things using AI.

1. Cleaned up the image
2. Upscaled the image to a higher resolution
3. Added stools and modified components
4. Changed the color palette to show different options

In the end, there is clarity in the image to show recommended enhancements.