We still benchmark models on design, but agents don't care about UI at all
tushaarmehtaa · x · 2026-09-08
tushaarmehtaa makes a contrarian point: new AI models are still implicitly benchmarked on design output — sexy landing pages, smooth animations, godly dashboards. Yet agents don't care about any of it. He says he hasn't properly opened the software he uses in weeks; ChatGPT work just goes in and does everything. Are we benchmarking models on building interfaces for a world where interfaces may no longer matter?
More from AGI Musings
- The Monobrain Problem: consulting firms collectively oversell half-baked agentic AI ideas to clients — DavidLinthicum · 2026-09-08
- As OpenAI/LLM rumored near Navier-Stokes breakthrough, global PISA math scores sink — IgorCarron · 2026-09-08
- "Is it AGI" flowchart from NeurIPS 2022 gets a call to re-test today's latest models — _rockt · 2026-09-08
- David Manheim: we've passed the singularity's 'front wall' as AI outpaces adaptation — davidmanheim · 2026-09-08
- Zvi Mowshowitz on Magic, Poker, Jane Street Trading and His AI p(doom) — TheZvi · 2026-09-08
- FT's Burn-Murdoch: digital distraction is eroding our capacity to focus and think — jburnmurdoch · 2026-09-08