Why AI Slop Is Making Design More Important, Not Less
August 16, 2026
Thank you!
Your submission has been received!
Your submission has been received!
Oops! Something went wrong while submitting the form.
Design is quietly becoming one of the most important roles in tech. Not because the tools got harder to use — they got easier. Everyone now has access to the same AI models, the same code generators, and the same component libraries, so the thing left to differentiate a product is judgment: what to build, what to leave out, and how it should feel to use.
No matter how fast you ship, AI-generated slop tells your users exactly what your product is: a cheap, vibe-coded wrapper. There's no faster way to break user trust. A generic layout, inconsistent spacing, copy that reads like it was written by nobody in particular — users clock it in seconds, even if they can't say why. They don't think "this was made with AI." They think "nobody made a decision here," and for a B2B buyer evaluating whether to trust you with their data, their workflow, or their team's time, that's disqualifying.
Speed was never the differentiator. Every team ships fast now. The differentiator is whether what you ship looks like someone made a decision about it.
That's not an argument against AI-generated interfaces. It's an argument about where they belong. Slop has one legitimate home, rapid prototyping, where the point is to test an idea, not sell it. When you're validating a hypothesis — does this flow make sense, will anyone use this feature, is this even the right problem — a rough, fast, disposable interface is exactly the right tool. Polishing a prototype before you know if the idea survives contact with a real user is wasted effort.
The mistake isn't using AI to prototype. It's shipping the prototype. Teams under pressure to move fast skip the step where a rough idea becomes a considered product, and take the vibe-coded version straight to production — then wonder why trial signups don't convert, or why enterprise buyers stall in procurement.
Once that idea is validated, opinionated design isn't a layer on top of the product. It IS the product. Not just the visual polish — the decisions about what the user sees first, what stays hidden until they need it, what the product refuses to let them do by accident. Those decisions are the product as much as the code underneath it. A tool with no opinion, one that shows the user every option at once because nobody decided what mattered most, isn't a finished product. It's a prototype with a production URL.
Use AI to throw ideas away fast. Don't confuse the speed of prototyping with the standard your production product needs to hit. The gap between the two is where design work actually happens.
If your roadmap process routes feature scope through product and engineering but treats layout, hierarchy, and interaction as an afterthought, you're building products with only two-thirds of the opinions they need. The teams shipping the best B2B SaaS products right now aren't the ones with the most features. They're the ones that said no to the most options.
An AI model can generate a screen, but it can't tell you what your product should refuse to do, or what your users actually need versus what they'll say they want. That's still a human call, made by someone who understands the problem, the users, and the business. AI can help you ship the opinion faster. It can't form the opinion for you.
Justin Roberts is the founder of Reloop Digital, a B2B SaaS consulting practice specialising in product strategy, UX design, and digital delivery. If your team is shipping fast but isn't sure the product feels considered, get in touch.