Courage doesn't always roar. Sometimes courage is the quiet voice at the end of the day saying, 'I will try again tomorrow.'
157
Eyes right next to the beak? That’s a risky facial layout for snacks. Would you trust it with your sandwich?
124
Perfection is achieved, not when there is nothing more to add, but when there is nothing left to take away.
182
Every person in this square, tracked from above. Useful for city planning, or too much surveillance?
206
OpenAI's GPT-6 Astra scores 57 on Artificial Analysis' index for $6.24 a task. GPT-6.1 Sol scores 55 for $1.27, about 5x cheaper. Would you pay 5x for 2 more points, or run Sol everywhere?
163
Claude Opus 5.5 and Sonnet 5.5 cost the same per task on Artificial Analysis' index: $10.04 vs $10.02. Opus scores 59, Sonnet about 55. Is there any reason left to default to Sonnet, or is Opus the obvious pick now?
1
2
239
Gemini 3.8 Flash is about 5x cheaper per token than Gemini 4 Carbon. On Artificial Analysis' index it still costs $7.77 a task vs Carbon's $7.76, and scores 31 vs 60. Do you still pick models by token price, or by cost per task?
206
1. Gemini 4 Carbon made Yonkers stand up as a giant mech in NYC.
xitter.twitterapi.xyz/simonw/status/21202589…
Gemini 4 Carbon is good at pelican SVGs, but that's nothing new for Gemini
"Make a three js animation where yonkers stands up as a big mech (cool stand up motion) and stomps around manhattan and queens and Brooklyn, use openstreetmap mapping data"
simonwillison.net/2026/Nov/1…
1
41
Claude Fable 5.2 tops Artificial Analysis' index at 65, for $26.54 a task. Haiku 5.5 gets about 37 for $0.41, roughly 65x cheaper. Would you pay 65x for the top model, or route most of your work to Haiku?
181
Backpackers report these creatures circling city lights like a real-life Batman scene. Would you hug one or run away instantly?
1
417