conv.

All stories
TechQuiet 9d · day 15

Nolan Lawson benchmarks how AI agents build the same app across JS frameworks

A new blog post has an AI coding agent rebuild an identical app in multiple JavaScript frameworks to test unproven "better for agents" claims.

newest coverage 9d agosummary written 10d ago55 pieces479 likes & upvotes388 comments Hacker NewsLobstersMastodonNewswires

What to know

  • Lawson had an AI agent (Opus 5.5) build an identical app across multiple JS frameworks, each using its official MCP/skills, to remove bias toward better-represented tools.
  • The benchmark is aimed at the web-dev community's pattern of making framework decisions on unverified "vibes" rather than evidence, now extended to AI-era claims that some frameworks suit agents better.
  • Lawson cautions that results are highly stochastic since LLM output varies between runs, limiting how definitive any single comparison can be.
  • The post needed several resubmissions on Hacker News and Lobsters over four days before gaining real traction on the HN front page.

Nolan Lawson Web developer and bloggerAlex Russell Web engineer cited in the postBaldur Bjarnason Commentator cited in the post

Nolan Lawson benchmarks how AI agents build the same app across JS frameworks
nolanlawson.com

How it unfolded 2 developments, newest first · click a bar or a number to jump articlesposts

Peak 12 pieces in 4h at Nov 1, 6 AM; 55 pieces over 15 days (2 articles · 8 posts · 45 comments) Oct 28, 7 AM — 2 pieces · 2 posts — Hacker News 2Oct 28, 11 AM — quietOct 28, 3 PM — quietOct 28, 7 PM — quietOct 28, 11 PM — quietOct 29, 3 AM — quietOct 29, 7 AM — quietOct 29, 11 AM — quietOct 29, 3 PM — 2 pieces · 2 posts — Hacker News 1, Lobsters 1Oct 29, 7 PM — quietOct 29, 11 PM — quietOct 30, 3 AM — quietOct 30, 7 AM — quietOct 30, 11 AM — quietOct 30, 3 PM — quietOct 30, 7 PM — quietOct 30, 11 PM — quietOct 31, 3 AM — quietOct 31, 7 AM — 1 piece · 1 post — Hacker News 1Oct 31, 11 AM — quietOct 31, 3 PM — quietOct 31, 7 PM — quietOct 31, 11 PM — 3 pieces · 2 articles · 1 post — Newswires 2, Hacker News 1Nov 1, 2 AM — 7 pieces · 2 posts · 5 comments — Hacker News 5, Mastodon 2Nov 1, 6 AM — 12 pieces · 12 comments — Hacker News 12Nov 1, 10 AM — 5 pieces · 5 comments — Hacker News 5Nov 1, 2 PM — 6 pieces · 6 comments — Hacker News 6Nov 1, 6 PM — 1 piece · 1 comment — Hacker News 1Nov 1, 10 PM — 2 pieces · 2 comments — Hacker News 2Nov 2, 2 AM — 4 pieces · 4 comments — Hacker News 4Nov 2, 6 AM — 6 pieces · 6 comments — Hacker News 6Nov 2, 10 AM — 3 pieces · 3 comments — Hacker News 3Nov 2, 2 PM — 1 piece · 1 comment — Hacker News 1Nov 2, 6 PM — quietNov 2, 10 PM — quietNov 3, 2 AM — quietNov 3, 6 AM — quietNov 3, 10 AM — quietNov 3, 2 PM — quietNov 3, 6 PM — quietNov 3, 10 PM — quietNov 4, 2 AM — quietNov 4, 6 AM — quietNov 4, 10 AM — quietNov 4, 2 PM — quietNov 4, 6 PM — quietNov 4, 10 PM — quietNov 5, 2 AM — quietNov 5, 6 AM — quietNov 5, 10 AM — quietNov 5, 2 PM — quietNov 5, 6 PM — quietNov 5, 10 PM — quietNov 6, 2 AM — quietNov 6, 6 AM — quietNov 6, 10 AM — quietNov 6, 2 PM — quietNov 6, 6 PM — quietNov 6, 10 PM — quietNov 7, 2 AM — quietNov 7, 6 AM — quietNov 7, 10 AM — quietNov 7, 2 PM — quietNov 7, 6 PM — quietNov 7, 10 PM — quietNov 8, 2 AM — quietNov 8, 6 AM — quietNov 8, 10 AM — quietNov 8, 2 PM — quietNov 8, 6 PM — quietNov 8, 10 PM — quietNov 9, 2 AM — quietNov 9, 6 AM — quietNov 9, 10 AM — quietNov 9, 2 PM — quietNov 9, 6 PM — quietNov 9, 10 PM — quietYesterday, 2 AM — quietYesterday, 6 AM — quietYesterday, 10 AM — quietYesterday, 2 PM — quietYesterday, 6 PM — quietYesterday, 10 PM — quietToday, 2 AM — quietToday, 6 AM — quietToday, 10 AM — quietToday, 2 PM — quietToday, 6 PM — quiet 12
Oct 29Oct 31Nov 2Nov 4Nov 6Nov 8now · 7:24 PM ET
  1. 2

    Benchmark post reaches Hacker News front page after repeated reposts

    After multiple earlier submission attempts on Hacker News and Lobsters between Oct 28 and Oct 31 drew little attention, a new submission by user luu reached the HN front page on Nov 1, picking up a higher score and comments, with bot accounts relaying it on Mastodon shortly after.

    “One thing I learned from this experiment is that everything I'm measuring is very stochastic.”
    — Nolan Lawson

    • AI cost of using anything is a good yard stick to measure things by. Red flags are a high degree of verbosity in tool outputs, complex failure modes, and slow running tools. Basically all that results in AIs having to look and dig through more output (tokens) and spend a lot of time polling tools for readiness (more tokens). The more complex a…

      jillesvangurpHacker News9d agoview on Hacker News ↗
    2 more of the top 3 · 45 posts in this stretch

    • IMO, the largest benefit of most JS frameworks is a great story for statee synchronization.Imagine an e-commerce website that has a cart with some products. The site has a cart button showing the current total, a pop-up cart interface, as well as an assortment of 10 or 20 different components which let you add or remove products (E.G. the…

      miki123211Hacker News10d agoview on Hacker News ↗

    • It's astonishing to me that even after a decade of SPA frameworks, most authors don't understand how a browser engine like Chrome works. There's enormous cost of manipulating the DOM from JS, as HTML elements get bound to JS objects lazily.An actual high performance framework should manipulate HTML as a string directly, or through HTML templates…

      torginusHacker News10d agoview on Hacker News ↗
    all of them →
  2. 3 days quiet
  3. 1

    Related piece warns AI coding erodes developer 'Fingerspitzengefühl'

    A separate Architecture Notes article argues AI coding tools like Cursor and Claude Code risk dulling developers' intuitive feel for code structure, citing a Reddit user's complaint about inconsistent, hard-to-trust AI output.

    “Where are the benchmarks?”
    — Nolan Lawson, Blog author · source
  4. background

    Lawson publishes AI-agent benchmark of JS frameworks — Nolan Lawson has an AI agent given the same spec build a searchable Pokémon data app in multiple JavaScript frameworks, each using its official MCP or skills, then tests the outputs with Lighthouse for performance and accessibility.

Also covered reported alongside — the timeline has no entry for these yet

  1. 2 outlets The AI cost of JavaScript frameworks

    first by HN Best, 10d ago · also HN Frontpage

What people are saying 21 voices from 1 site · best of 45 · verbatim