AI for Interior Designers™ Podcast

Confessions of AI Designer Nerds: Behind the Screens & Scenes

A candid roundtable with fellow designer-technologists on what is actually working, what got cancelled, and why the designers talking most negatively about AI are usually the ones using it without knowing it.

Ep. 59 ✦ Jenna Gaidusek ✦ Jessica Nelson & Stephanie Lindsey Tools Mentioned September 2025

This blog was written using AI as a recap from the recording, then edited by the author for accuracy and details.

Key Takeaways

Start with an LLM and let it teach you to prompt. Pick one — ChatGPT, Gemini, Claude — and that skill transfers to every other tool you add later.
The biggest AI wins are on the back end of the business, not the visual side. Automating reports, eliminating triple-entry procurement, streamlining onboarding — this is where hours actually come back.
Clients are impressed, not alarmed, when you tell them you use AI. Framing matters: AI is the ideation tool that gets their vision on screen faster. The design judgment is still yours.
Nano Banana keeps images consistent. ChatGPT changes things you told it to leave alone. For precision product placement, Gemini's image tools are currently winning.
The designers most vocal against AI are usually already using it. Clients send AI-drafted emails. Designers use AI-powered tools without calling them that. Transparency is worth embracing openly.

Jessica Nelson and Stephanie Lindsey – Etch Design Group

Episode Guests

Jessica Nelson & Stephanie Lindsey

Co-Founders, Etch Design Group & AI for Interiors

Jessica and Stephanie are the co-founders of Etch Design Group, a full-service interior design firm, and AI for Interiors, an education and advisory platform helping designers integrate AI. They also co-host the AI for Interiors Podcast and 100 Lunches Podcast. This episode was part of a cross-podcast collaboration — they also interviewed Jenna on both of their shows.

Interior Design AI Education Automations Business Systems Podcast Hosts

Etch Design Group ↗ AI for Interiors ↗

The Short Answer

Three designers who are actively building, advising, and using AI sat down and compared notes. The conversation is less about which specific tools to use and more about the shift in how you think about your business — from identifying the tasks that drain you, to building automations that handle them, to getting comfortable showing clients the process behind the visuals.

How Designers Are Actually Entering the World of AI

The entry point for most designers is an LLM — ChatGPT, Gemini, or Claude. Not image generation. Not video. Just conversation. Jessica's advice: pick one, get comfortable talking to it, and let it help you figure out how to prompt the other tools. Once you understand how to communicate with a chat-based AI, that skill transfers everywhere.

The typical progression is toward image tools and then video — but the designers getting the most value are the ones who went a different direction entirely: toward the back end of their business.

"Designers feel more comfortable starting with more mundane task-type assistants versus the design process — because the design process is their baby and they know the creative side."

— Stephanie Lindsey

Custom GPTs for drafting budgets on client calls. Automations that eliminate triple-entry procurement workflows. Email assistants that draft responses in your voice. These are not glamorous, but they are where hours come back every week.

The Back End Is Where the Real Time Is

Jessica and Stephanie have been systematically identifying and automating the repetitive bottlenecks in Etch's operations. The most vivid example from the episode: a procurement coordinator at a firm they were onboarding who manually entered the same information into three separate programs — Studio Designer, Asana, and a client report — every single time. Identical data, three times over. One automation eliminates that entirely.

The question to ask yourself: what do you do in your week that you could describe as "entering the same information into two different places"? That is an automation waiting to be built — and you likely do not need a developer to build it.

Jenna shared her own version: a team app she built on a Friday that automatically scrapes new blog posts and videos, generates a brand-voice summary, and populates a weekly newsletter with one button click. The app assembles everything; Jenna reviews it before it sends. That is the right human checkpoint — enough automation to save hours, enough oversight to catch anything that is off.

On Visualization Tools: What Works, What Does Not

Jessica and Stephanie use SketchUp and AutoCAD as their technical foundation and bring AI in at the rendering layer. Their current stack for visualization:

Arch Synth — converts product images to 3D models for import into SketchUp. If you want to show a specific selected piece inside the actual model, this is the tool.

ProAI (PROME) — takes a SketchUp screenshot and renders it. Two modes: sketch rendering (prompt-guided, gives three options per render) and texture lock rendering (materials-applied, no prompt needed). A creativity slider controls adherence — they keep it at 20 or below.

Nano Banana (Gemini) — praised by all three for keeping images consistent. Makes what you ask for and does not touch anything else. Currently the standout for room-level precision editing.

"Nano Banana is like: I'm going to keep everything the same and only do what you want me to do. ChatGPT is like: I wasn't going to touch the window, but I did anyway."

— Jessica Nelson

The broader note: ChatGPT is good for material swaps on individual items — changing a fabric on a single chair, for instance. It struggles with room-level edits where you need it to change one thing and leave everything else exactly as is. Nano Banana handles the latter far more reliably right now.

Telling Clients You Use AI — and Why It Lands Better Than You Think

Stephanie and Jessica's approach: they tell clients upfront in the first inspiration meeting that they will be seeing AI-generated images and renderings. Not buried in a contract — said out loud in the first conversation. Clients, 100% of the time, respond positively. The most common reaction is "that is amazing, I love that you are using every tool available."

The framing that works: AI is not designing for you. It is helping you visualize what is already in the client's head so you can confirm direction faster. The design judgment, material curation, and real-world execution are still entirely yours.

The concern that clients will be put off is mostly designers projecting their own ambivalence. The clients who hire professionals are not worried that you are using tools efficiently. They are worried you will not understand what they want. AI helps you show them that you do — faster and more convincingly than a concept board that requires mental translation.

On Ethics, Copying, and Why the Honest Designers Will Win

The conversation went somewhere worth capturing. The designers most vocal against AI are frequently using it. Clients are sending AI-drafted emails. Designers are using image tools with AI embedded and not calling it that. The line is already blurrier than the public debate acknowledges.

Stephanie made the observation that copying was happening long before AI existed — she described a local designer who had replicated one of Etch's powder rooms almost identically using reverse image search and Amazon alternatives. Not AI. A lack of creativity combined with access to reference images. AI makes the detection more visible, not the behavior more common.

Where AI changes the ethics calculus: there is no longer any resource constraint that justifies using another designer's project images in your own proposals. You have the tools to generate original concept imagery from your client's own words. Using someone else's work when you have the capability to create your own is a choice, not a necessity.

"If you're not generating your own images for proposals, what are you doing? You have the tools. Stop stealing people's work. Visually communicate what you want to see."

— Jenna Gaidusek

Frequently Asked Questions

Pick one large language model — ChatGPT, Gemini, or Claude — and start using it for things you already do: drafting emails, answering client questions, researching products. Let it help you write a prompt for an image tool when you are ready. Once you are comfortable communicating with one AI in conversation, that skill transfers to every other tool. The biggest barrier is starting. The tool will teach you how to use it if you let it.

The highest-impact applications: custom GPTs for drafting client budgets on discovery calls, automations that eliminate data re-entry across platforms like Studio Designer and Asana, email assistants that draft in your voice, onboarding workflows that customize your approach to each client personality type, and weekly report generators. These save more compounding time than image generation does for most firms — and they scale in value as your business grows.

Say it plainly in the first conversation — not in a contract footnote. "In your inspiration meeting, you are going to see AI-generated images. We use these to quickly visualize what you are describing so we can confirm direction before committing to materials." Most clients respond positively. The framing that works: AI is the ideation tool that gets your vision onto a screen faster. The design decisions, curation, and execution are still entirely yours.

ProAI (spelled PROME) is a rendering tool that takes a screenshot of a SketchUp model and renders it using AI. It offers two modes: sketch rendering, where you provide prompts and get three image options at a time, and texture lock rendering, where it renders the materials you have already applied without requiring a prompt. A creativity slider controls adherence to the original — keeping it at 20 or below maintains accuracy while adding realistic material quality to the output.

Nano Banana (Google Gemini's image editing tool) follows precision instructions reliably — "add this product here and change nothing else" — and actually does that. ChatGPT tends to alter things you explicitly told it to leave alone, which is frustrating when you have a composition you like and just need one element swapped. For room-level edits where precision matters, Nano Banana is currently the more reliable choice. ChatGPT can be useful for material swaps on individual items, but struggles with whole-room edits.

Cross-Podcast Collaboration

Also Listen: AI for Interiors Podcast & 100 Lunches

This episode was recorded alongside two reciprocal interviews — Jessica and Stephanie also hosted Jenna on the AI for Interiors Podcast and 100 Lunches Podcast. Two very different conversations worth hearing from both sides.

AI for Interiors ↗ Certificate Program ↗