Is it still worth learning UI/UX as a Beginner?
Chapters6
Evaluates latest frontier models like Gemini and Codex, rating UI UX quality around 6/10 and coding stronger around 7-8/10.
AI can assist UI/UX work, but beginners still need eye for detail; humans outpace AI now, especially on fine-tuning and visual polish.
Summary
DesignCourse’s latest quick take dives into how current frontier AI models perform in UI/UX versus coding. The creator rates UI/UX capability at about 6/10 and code capability higher, around 7–8/10, using models like Gemini, Codex, and Opus as benchmarks. He emphasizes that, despite AI’s output, a trained designer can spot serious UX and typography issues that AI often misses, citing real-world project examples from vibe-coded designs shared on Reddit and Facebook groups. The discussion includes concrete critique of several sample designs—from logo symbolism and spacing to inconsistent color schemes and typography—to illustrate how AI-generated designs still need human refinement. He argues that beginners shouldn’t rely solely on AI; instead they should use AI as a macro-design tool to accelerate workflows while developing a trained eye to critique and improve AI output. The host reiterates that designers aren’t cooked yet: with hard work, entry-level designers can surpass the AI six out of ten and grow beyond it. Finally, he shares his practical approach: use AI for rapid ideation and layout, then apply meticulous, human-level fine-tuning at the end of a project. He finishes by inviting viewers to subscribe and engage, underscoring that AI will augment, not replace, core UI/UX skill growth for years to come.
Key Takeaways
- Frontier AI models (Gemini, Codex, Opus) rate around 6/10 for UI/UX and 7–8/10 for code, according to the creator.
- Human designers can still spot critical UI/UX issues in AI-produced work, such as logo symbolism, spacing, alignment, and color scheme mismatches.
- Beginner designers should use AI as a tool to rapidly build and prototype, while actively developing the eye to critique and improve AI output.
- Despite AI progress, there is no imminent replacement for skilled UI/UX designers; entry-level designers can still outperform AI with effort and practice.
- The creator intentionally blends AI-assisted workflows with traditional polish: leverage AI for macro design, then apply fine-tuning and typography refinement manually.
- Real-world critiques from design communities (Reddit, Facebook vibe coding groups) illustrate consistent UX/UI problems that AI often cannot resolve on its own.
- Committing to ongoing learning and hands-on practice remains essential for staying competitive as AI capabilities mature.
Who Is This For?
Aspiring UI/UX designers wondering if learning design is worth it when AI can generate layouts; this video argues that AI is a tool, not a replacement, and that dedicated practice still wins in the long run.
Notable Quotes
"When it comes to UI UX, it's probably around a six out of 10."
—Initial rating of AI’s UI/UX capabilities relative to human designers.
"I use it almost maximally as possible, but I'm also refining it."
—Describes his practical use of AI as a high-speed helper with manual refinement at the end.
"Are designers cooked or whatever? No."
—Reassures beginners that AI won’t fully replace designers yet.
"You still need to try to develop the eye so that you can spot this stuff."
—Emphasizes the ongoing need for human critique skills.
"If you're entry-level and you're a beginner, well, you're still not necessarily cooked."
—Encourages beginners that they can grow beyond AI with effort.
Questions This Video Answers
- Is UI/UX still worth learning when AI can generate designs quickly?
- How does AI compare to human UI/UX designers in 2026?
- What should beginners focus on to outrun AI in UI/UX design?
- Can AI-assisted design workflows replace typography and spacing refinements?
- What are the best practices for combining AI tools with traditional UI/UX critique?
UI/UX DesignAI in UI/UXFrontier AI modelsGeminiCodexOpusVibe codingTypographyUI critiqueDesignCourse
Full Transcript
So, just how good are the latest frontier models in terms of UI UX and code? So, I gave this a thought as somebody who's really dived completely in in the last year and a half or two into AI coding and some design as well. And this is what I've come up with. I think right now with the latest frontier models, what we're talking about, you know, Gemini, Codex, Opus, etc. When it comes to UI UX, it's probably around a six out of 10. Zero being the biggest noob ever and then [music] 10 being the absolute best UI UX designer.
Now, of course, when it comes to code, I personally think it's stronger in the code domain, so I put its price somewhere between seven and eight out of 10. But since I'm a designer of multiple decades and first and foremost my skill, my best skill is UI UX, I think I have a really good read on this. So, take a look at some of these projects for instance. I went on Reddit and I went on Facebook and these vibe coding groups and people are sharing their projects openly. And I just kind of wanted to show, I you know, what these vibe designed looking projects look like.
So, here's one right here. Okay. Um as a UI UX designer, I would take a look at this immediately and say, "Okay, there's an issue right here." Like this this this V, this symbolism, first from a UX perspective, it's unnecessary just to have this huge logo. From a UI perspective, it's way too close to this stuff right here. And there's also some spatial issues as well and alignment issues. And this is something, you know, somebody who's very trained will be able to spot. Now, you ask the AI to produce something like this, it will certainly produce it and you can tell it to be the best UI UX designer ever and it will still produce things that are inadequate.
Uh how about this one? Okay. Again, you know, we're on you know, desktop, why not utilize more of the desktop? Okay, maybe it's a mobile only approach. Okay, well, let's dig in there. Well, look at the colors, you know, we have three different uh color schemes occurring here. The stroke on them doesn't even match with this green or this blue next to like this this burnt orange color. There's a lot of things like this I in type typography as well that I can just sit here and spot out. I used to do these reviews where I would just refactor people's designs and you know, like somebody, you know, who's been doing it for decades can really spot the issues like crazy.
Like here's another one. I Again, you know, this type of thing I I I'll I'll rate this like a four out of 10, five out of 10. Um how about this one? Yeah, not a fan of the colors here. Gem Hunter, I it's it's not really telling me exactly what's happening here. This is kind of like a poor hero section. Uh the cards definitely need work with the typography in here. Um War Room Academy, again, like like we can do way way better than this. Like a really skilled UI/UX designer. Here's another one. I this one screams of, you know, typical vibe coded design.
I nothing special happening here. So, my point is is well, there's multiple points really. Um if you're somebody who's starting out in UI/UX design, I obviously you're probably not going to be as good as what the current LLMs can produce. So, in that case, you use it as a tool to help you build instead of you just trying to raw dog it yourself. But at the same time, simultaneously, you still need to try to develop the eye so that you can spot this stuff and eventually get better than the AI itself. So, when people ask the question, are you know, is are designers cooked or whatever, at at the current time, no.
If you're entry-level and you're a beginner, well, you're still not necessarily cooked. It all depends if you're you're willing to do the old school approach, which is learn stuff, you know, like apply yourself, try to get better. I mean, you can still be somebody who's brand new and enter this industry as long as you're willing to work hard, you can get better than this six out of 10. Of course, you can. Now, one last point I'm going to make in this quick video is yeah, I consider myself way better than AI in terms of UI UX design, but does that mean I don't use it?
No, I use it. I I use it almost maximally as possible, but I'm also refining it. But I'm using it kind of like a entry-level designer that can knock out a lot and flush out things really quickly from like a macro perspective, a high-level perspective, and then at the end of a project when I have all the functionality and stuff working, then that's when I go in and I really start making fine-tune adjustments to the UI. So, that's the sort of thing that, you know, I'm focused on this channel is just showing that there's still a great amount of skill involved in using AI for both design and code, and I would say more so with design.
So, make sure to subscribe up here, check out my courses at designcourse.com. That's the core focus of what this channel's about is really trying to maximize the use of AI, but also our specific skill sets because trust me, we're we're not going to get to a point in time where AI is just better than everybody, no matter what. That's years in the distance. So, go ahead, take that for what you will. Let me know what you think in the comments. Do you think I'm being too generous with how good AI is or is it even better than what I'm saying?
I'll see you all soon, and goodbye.
More from DesignCourse
Get daily recaps from
DesignCourse
AI-powered summaries delivered to your inbox. Save hours every week while staying fully informed.









