Show HN submissions tripled and are now mostly vibe-coded

HN’s front page is glowing purple — lazy “vibe-coded” slop or just shipping faster

TLDR: Show HN posts have surged, prompting limits on new accounts, and an analysis claims many projects share “AI-looking” design patterns. The comments split: some see lazy, look‑alike pages; others say faster shipping and wider access matter more, and the upvote filter keeps the good stuff visible.

Show HN is having a glow-up — literally. Submissions have tripled, enough that moderators clamped down on new accounts. One poster scanned 500 project pages for telltale “AI design” — think Inter font, purple gradients, glassy cards — and found 21% heavy slop, 46% mild, 33% clean. The method? A scripted browser checked page styles; no judgment calls, just pattern-spotting.

But the code isn’t the drama — the comments are. Is purple glow the new scarlet letter, or just a shortcut to ship? julia-kafarska drew a line in the sand: there’s a difference between a “vibe-coder” and an engineer using AI to work faster. simonw fired back that most side projects are AI-assisted now and even nitpicked the headline’s use of “vibe-coded.” Meanwhile, fooker shrugged: if it makes the front page, the upvotes did their job. More cheerleading came from cr125rider: better tools mean more people can hack on problems. And then came the meme moment: cammasmith confessed they can’t unsee the patterns anymore — the “VibeCode Purple” and colored borders feel like the em-dash of design.

Translation: half the crowd is clutching pearls over copy-and-paste gradient cards, the other half says ship it and let the votes decide. Font police vs. shipping club — pass the popcorn.

Key Points

  • Show HN submissions have significantly increased in recent years (described as tripled), prompting HN moderators to restrict new-account submissions.
  • The author compiled a list of recurring AI-influenced design patterns across fonts, colors, layout, and CSS styles.
  • 500 recent Show HN landing pages were programmatically analyzed using Playwright and deterministic CSS/DOM checks without screenshots or LLMs.
  • Manual QA estimates a 5–10% false positive rate; the author may open-source the scoring code for replication.
  • Results: 21% of sites triggered 5+ patterns (“heavy”), 46% triggered 2–4 (“mild”), and 33% triggered 0–1 (“clean”).

Hottest takes

"There's a big difference between vibe-coder and engineer who uses ai to speed up their work." — julia-kafarska
"the voting mechanism is working as a good filter." — fooker
"I expect most side-projects are being built with AI-assistance now." — simonw
Made with <3 by @siedrix and @shesho from CDMX. Powered by Forge&Hive.