The Design Episode: UI/UX in the Age of AI & Vibe Coding | Attempted Intelligence Ep. 5

Nov 11, 2026

When AI can spin up a polished prototype in an afternoon, what happens to the traditional design process—and where does that leave human designers?

In Episode 5 of Attempted Intelligence, hosts Harlan Parrott (VP of AI Innovation at KnowBe4) and Matt Duren (SVP of Engineering for AI & Data) sit down with Anandi Ellis, Design Architect at KnowBe4. Anandi traces her journey from graphic and label design to UX/UI and software engineering, sharing firsthand how generative AI tools have upended the way digital products are conceptualized, prototyped, and built.

Together, they explore why the conventional design workflow is nearly obsolete, what happened during Anandi's week without Figma, why pen-and-paper ideation is making a comeback, and why taste, curation, and expert polish matter far more when anyone can generate a "finished-looking" interface in minutes.

KEY TAKEAWAYS FROM EPISODE 5:

The Death of Slow Prototyping: How AI compressed design cycles from weeks to hours—and how designers must adapt.
The Value of Human Taste: Why non-experts mistake quick AI generation for quality, and why distinguishing "good" from "great" requires deep craft.
Vibe Coding vs. Production Code: What happens when engineers inspect AI-generated front ends, and the balance between rapid ideation and maintainable architecture.
Generalists vs. Specialists: How tools like Claude skills and AI coding assistants are reshaping cross-functional team roles.

CHAPTERS / TIMESTAMPS:

00:00 - Introduction to Attempted Intelligence

00:55 - Welcome to Episode 5: Meet Anandi Ellis

01:26 - Anandi’s Path From Graphic Design to KnowBe4

04:43 - Moving Beyond the “Hit It With a Pretty Stick” Mindset

07:49 - Why the Traditional Design Process Is Almost Dead

09:06 - Experimenting with AI: Anandi’s Week Without Figma

12:35 - Back to Pen & Paper: Deciding What Deserves to Be Built

14:14 - Harlan’s Purple Website & Spotting AI Slop

16:49 - “That’s So AI”: What a Six-Year-Old Understands About Taste

19:34 - The IKEA Question: When Is AI Design "Good Enough"?

22:53 - Vibe Coding: What AI Front Ends Look Like to an Engineer

24:39 - Scaling Knowledge: Sharing Expertise Through Claude Skills

35:41 - The Future Workforce: Generalists vs. Specialists

FEATURED QUOTES:
💬 “If you’re not using AI to do your job better and faster, take that challenge. Actually go and see if you can completely disrupt the way you are currently working.” — Anandi Ellis
💬 “If you’re starting to use AI for something that you don’t know that much about, then what you end up with is an output that seems to be very high quality, but then to an expert is actually very low quality.” — Matt Duren
💬 “Everybody can do something relatively quickly and cheaply. But the question is also, should you be doing it? Focus your time and effort on building the things that will make big impact.” — Anandi Ellis

RESOURCES & LINKS: