UnTangled with Taylor Wilson – AI Beyond the Hype: What’s Actually Working Right Now

There is a specific kind of exhaustion that sets in when you have heard too many AI takes. The doom scroll of headlines telling you everyone’s job is gone in 18 months. The Instagram ads promising a seven-figure business if you just download this ebook. It is a lot. And none of it is particularly useful if you are just trying to figure out how to do your actual job a little better.

That is exactly why I loved last week’s conversation on UnTangled with Taylor Wilson, founder and Chief Shizzle Officer of Creative Shizzle, a branding and creative marketing partner for nonprofits and mission-driven companies. Taylor has been in the marketing trenches for 20 years, she has been experimenting hands-on with AI tools inside her own business, and she has zero interest in hype in either direction. What she does have is a very practical point of view on where AI is delivering real value right now and where it is still mostly noise.

Here is what I took away.

The Hype Is the Problem, Not the Technology

Taylor named something I have been thinking about too. The loudest AI voices are creating two groups of people: those who are paralyzed by fear and those who are overly hopeful that the right tool will finally solve the big messy problems without much heavy lifting on their end. Neither of those is a useful place to be.

The hype that says AI will do everything for you is just as unhelpful as the hype that says AI will take everything from you.

What is actually true right now is somewhere in the middle. AI can remove friction from repetitive tasks. It can help teams move faster and think more clearly. But there is still no such thing as a free lunch here. Every tool requires real adoption effort, learning curves, and ongoing human judgment to get value out of it. Less work than before, maybe, but still work. The organizations that are getting the most out of AI are the ones that went in with realistic expectations and put in the time to actually learn the tools.

Where AI Is Actually Saving Time

Taylor walked through some real use cases from inside her business, and they are the kind of practical stuff that does not always make it into the headlines.

Answer engine optimization (AEO). Search has changed. People are typing full questions into Gemini, ChatGPT, Perplexity, and Claude instead of just Googling keywords. Tools that help organizations understand what their audiences are asking in these spaces, and how to show up in those results, are becoming a genuine strategic asset for nonprofits. Taylor has been using Adaptify for this and the insight it surfaces in minutes would take a team member weeks to compile manually.

AI inside tools you already use. Monday.com, MailChimp, your CRM, QuickBooks. Almost every platform nonprofits are already paying for is quietly rolling out AI features. One of Taylor’s best tips: instead of hunting for brand new tools, go back through the integrations pages of the platforms you already use. You will find things that were not there six months ago. This is one of my big recommendations too. Stop, take stock of what you already have, and give someone on your team the explicit job of doing that exploration before you spend a dollar on anything new.

The meeting notes to presentation workflow. Here is a simple sequence that Taylor laid out that you could implement tomorrow. Use an AI note taker in your meetings. Drop the transcript into ChatGPT or Claude and ask it to brainstorm additional ideas, pressure test the thinking, or pull out the key themes. Feed that into Gamma to generate a clean presentation. Taylor’s estimate for this whole process is about an hour. If you have ever spent a week getting a leadership presentation together, you know what that kind of time savings actually means.

Build vs. Buy: Taylor’s Actual Take

Taylor’s position on this is nuanced, and I think she gets it right. For most organizations, the answer is still: buy the tools and work with what exists. But that does not mean building is off the table entirely.

Where building makes sense right now is in prototypes. Small, specific, low-stakes experiments that help you understand your own data or visualize something that existing tools cannot show you cleanly. Taylor gave a great example of feeding email campaign data into a vibe coding tool to generate a dashboard that made the top-performing campaigns immediately visible. No developer required. No massive build. Just a fast, useful prototype that helped a team make a decision.

The caution is real, though. Trying to build your own CRM or a full production application is a whole different thing. It becomes a part-time job fast, and you still need people who can think architecturally about systems even if you are not writing the code yourself. As my guest Tommy Spann from UpLift Partners put it, he rebuilt his entire platform stack and it consumed him for months. Worth it for him. Not the right call for most nonprofit teams.

The short version: buy what exists, prototype what doesn’t, and be honest with yourself about how much bandwidth you actually have.

The Harder Parts Nobody Is Talking About

AI tools are getting faster and more capable, but the adoption friction inside organizations is still real. Quality control takes time and attention. Prompting well is genuinely a skill that takes practice. Getting non-technical staff comfortable with new workflows requires training and patience. None of that disappears just because the technology is impressive.

One of the things I keep coming back to after conversations like this one is that AI works best when humans stay firmly in the loop at the interpretation layer. You can generate a first draft, a brainstorm, a dashboard, a presentation outline. But someone still has to evaluate whether it is actually right, relevant, and worth acting on. That is not a limitation to work around. That is the job.

A Quick Note on Accountability

Taylor’s meeting notes workflow is genuinely useful and I have been doing versions of it myself. But it is worth sitting with a question it raises: when an AI-assisted presentation shapes a strategic decision, who owns the quality of that thinking? Clicking approve is not quite the same as doing the reasoning. The more we use these tools to speed up the front end of our work, the more deliberate we need to be about who is actually accountable for what comes out the other side.

This is not a reason to avoid the tools. It is a reason to stay awake while you use them.

The Bottom Line

We are past the breathless early days of AI hype and into something more interesting: the part where organizations with real missions, limited budgets, and actual work to do start figuring out what is genuinely useful. Taylor is doing that work inside Creative Shizzle every day, and this conversation was a good reminder that the most valuable AI insight usually comes not from the headlines but from people who are just out here experimenting.

Watch the full episode to see the specific tools Taylor recommends and hear her talk through the workflows in her own words.

Watch the full episode here: 

Tools Taylor Mentioned

AEO / AI Search Visibility

– Adaptify (adaptify.ai) — deep AI visibility reports showing what audiences are asking in answer engines

– SEMrush, Moz, Search Atlas, Ubersuggest — traditional SEO platforms now adding AI visibility layers

Research and Strategy

– Perplexity — answer engine for research and discovery

Project Management

– Monday.com — now includes AI agents that build project boards from plain-language prompts

Content and Presentations

– Gamma (gamma.app) — turns outlines and notes into polished presentations

– Google NotebookLM — synthesizes and organizes research from multiple sources

– ChatGPT / Claude / Gemini — for brainstorming, synthesis, drafting, and pushing back on your own thinking

Vibe Coding / Prototyping

– Lovable — vibe coding tool for fast prototypes and dashboards

– Replit — browser-based coding and prototyping environment

– Base44 — no-code app builder

AI Note Takers

– Multiple options exist (Otter.ai, Fathom, Fireflies, etc.) — Taylor uses one to capture meetings and feed transcripts into downstream workflows

Tool Discovery Tip From Taylor

Go to the integrations pages of platforms you already use (MailChimp, HubSpot, QuickBooks, your CRM) and start clicking on tools you don’t recognize. This is one of the best low-cost ways to find things that are already compatible with your existing stack