Allie Miller: Claude Opus 5.5 is the first to pass her code-meets-poetry test
alliekmiller · x · 2026-09-24
Allie Miller long considered OpenAI models best for ideation — until Opus 5.5. It's the first model to pass her private "code meets poetry" challenge, which she'd shared with only one lab. In a slide-design test, Codex floundered while Opus 5.5 generated six labeled PNG options via GPT's image 2.5, saved them, opened Finder, stated its preferences, then auto-built her picks as fully editable PPT slides unprompted. She notes earlier tool-call failures appear fixed and calls it a tool-wielding machine, far more creative than prior Claude models.
More from Models
- Artificial Analysis launches TTS leaderboard with new Pronunciation Robustness Benchmark across 95 models — ArtificialAnlys · 2026-09-24
- Human text flagged 100% AI: why people treat stochastic detectors as oracles — tokenbender · 2026-09-24
- GPT-6 Sol and Luna Score Below GPT-5.6: First Major Release Weaker Than Its Predecessor — srchvrs · 2026-09-24
- Reddit users slam GPT-6 Sol and Luna as dumber than 5.6 despite cheaper pricing — Obvious_Resort8887 · 2026-09-24
- Opus 5.5 Is 10x Slower Than Fable for Financial Modeling, User Reports — JOBhakdi · 2026-09-24
- ChatGPT Voice Adds Email, Calendar and Slack Plugins, Powered by GPT-6 Astra — OpenAI · 2026-09-24