Ep 51: Simple Ways to Start Implementing AI and What’s Coming Next
Listen to the Podcast Episode for a deeper dive
Simple Ways to Start Implementing AI — and What's Coming Next
The tools you should be using right now, the workflow that takes meeting notes all the way to a polished proposal, and a clear-eyed look at what agentic AI will eliminate — and what it will never replace.
- If you are not using a notetaker, you are working ten times harder than you need to. Recording meetings and running transcripts through an LLM is the single highest-impact, lowest-barrier change you can make right now.
- The transcript-to-proposal workflow already works and saves hours. Take your meeting recording → extract the transcript → run through a custom GPT trained on your proposal format → pop into Canva. Done in minutes, not hours.
- AI can generate concept visuals from your client's actual words. Instead of scrolling Pinterest for another designer's work, prompt an image tool using the exact language your client used to describe how they want their space to feel.
- Beware AI-washing. Many tools calling themselves AI are just automations with new branding. Ask whether the tool does something your core LLM cannot already do for $20/month — and avoid annual subscriptions on anything that could be wiped out by a ChatGPT update.
- Agentic AI will take out the middleman steps. The multi-app juggling act — download from here, upload to there, reformat, move it over — is being eliminated. Build familiarity with today's tools so you are ready when that shift arrives.
Start Here: Why Notetakers Are Still the Most Underused Tool
Jenna still does presentations where people in the room do not know that meeting notetakers exist. Two years in, recording meetings and generating transcripts remains the highest-impact, lowest-difficulty change most designers can make immediately — and it is not even really an AI thing. The AI comes next. First, just capture what was said.
For virtual meetings: Loom now integrates directly with Zoom. It records the meeting, generates a timestamped outline, and sends the recap to everyone in the meeting automatically. No separate tool, no extra subscription if you are already using Loom for videos. Fathom is also solid for this use case.
For in-person meetings and job site walkthroughs: Jenna uses Plaud — a physical recorder (also available as a pin-style wearable) that picks up two-way conversations and generates searchable transcripts via the Plaud app. It can also attach to your phone to record phone calls when toggled on. Both iOS and Android also have native transcription tools in their notes apps that can handle voice recording without any additional tool.
The practical point: when you stop writing notes during a meeting, you become fully present with your client. You see the room. You catch the subtext. You pick up on the things that nobody writes down because nobody notices them while they are writing. The notetaker handles the capture so you can be the designer.
For hand notetakers who do not want to give up pen and paper: photograph your written notes and upload the image to ChatGPT or another LLM. Ask it to extract the text into typed form. Your handwritten notes become a searchable, copy-pasteable document in seconds — as long as your handwriting is legible enough for the model to read.
The Transcript-to-Proposal Workflow That Already Works
This is the most repeatable, immediately actionable workflow in this episode. Once you have a meeting transcript, the path from that conversation to a polished, branded proposal is a series of steps that AI handles most of for you.
"I still write proposals, but I don't write them. I say the things and then have AI repurpose them in a way that I like. Now it's smooth."
— Jenna GaidusekThe time difference is significant. What used to take a few hours — gathering, writing, formatting, making sure nothing was missed — now takes a fraction of that. More importantly, nothing gets lost. The transcript captures every detail the client mentioned. The AI organizes it. You make sure it is right before it goes out.
Going Further: Task Lists, Deadlines, and Calendar Imports
For designers who want to take the transcript workflow a step further, AI can also turn meeting notes directly into assignable task lists with deadlines — and export them in formats your calendar and project management tools can import.
CSV stands for comma-separated values — essentially a text file that spreadsheet programs and calendars can read as structured data. ChatGPT can generate one correctly formatted for your specific calendar if you tell it which platform you use.
How to Spot AI-Washing — and Avoid Expensive Mistakes
Jenna draws a direct parallel to greenwashing in the design industry: when sustainability became a buzzword, companies slapped "eco-friendly" on products that had only marginal environmental improvements to capture market interest. The same pattern is happening now with AI. Many tools are labeling themselves AI when they are running simple automations with no actual language model behind them. Others are genuine AI tools — but built on ChatGPT or Stable Diffusion APIs, which means a major ChatGPT update can effectively wipe them out overnight.
She points to what happened when ChatGPT's DALL-E image generator got a major upgrade in late March 2025: several stable diffusion-based image tools that charged $60+ per month for the same capability were suddenly obsolete. The parent technology was now available to everyone for $20.
- Does this solve a real, specific problem in my workflow that my existing tools cannot?
- Is it built on a foundation model (ChatGPT, Gemini, Claude) — and if so, could a direct update to that model eliminate what makes this tool special?
- Has this company been around long enough to have a track record, or is it a new startup riding the AI wave?
- Am I being asked to commit to an annual subscription? If so, why — and is there a legitimate reason this tool will be around in a year?
Jenna's subscription rule: pay annually only for foundational models (ChatGPT, Gemini, Claude, Perplexity) or for tools deeply integrated into your existing workflow software. For newer, single-purpose AI apps, stick to monthly until the tool proves it has staying power and does something the foundation models cannot replicate.
What Agentic AI Changes — and What It Does Not
The honest framing of where we are right now: we are still doing a lot of manual handoffs between tools. Copy from here, paste into there, download this, upload that. The next phase — agentic AI — takes out those middleman steps by having AI execute the full sequence rather than just each individual step when prompted.
- Record meeting → manually copy transcript
- Paste into ChatGPT → generate proposal
- Download → reformat → paste into Canva
- Generate images separately → add to template
- Export PDF → manually send email
- Upload conversation → AI handles the entire sequence
- Proposal, images, branded PDF — generated automatically
- Email drafted and queued for your review
- You check it, approve, send — nothing more
- All the steps in between: gone
Jenna is careful to frame this as exciting and concerning at the same time. The monotonous, time-wasting tasks that have eaten designers' hours for years will be eliminated. So will some entry-level roles that were built around doing those tasks. The designers who will be most protected are the ones who can do what AI will never do: tell the human story, bring genuine taste, read the room, understand how a space needs to function for a specific person's life. That is not in training data. It is in experience.
"Don't ask for something that eliminates your job. Ask for things that make your job easier so you can get back to the parts that make you you."
— Jenna GaidusekJenna is the go-to educator for design professionals who want to use technology without losing their creative edge. A designer turned tech advocate, she's a nationally recognized speaker, podcast host, community builder, and custom app builder based in Charleston, SC.
Disclaimer: This blog was written using AI as a recap from the recording then edited by the author for accuracy and details.
