Every Frontier team: one headline rewrite lifted agent citations from 2/240 to 240/240
every · x · 2026-10-29
Every's Frontier team published its weekly log of AI experiments:
- Rewriting a single headline took simulated agent citations from 2/240 sessions to 240/240
- Testing whether agents wait on slow APIs: one agent waited 30 seconds in all 20 trials, but most never tried
- Using LLM neurons to make memes measurable and trace their family tree
- Having AI anticipate Dan Shipper's takes: 400 guesses covered 93%, though picking the right guess remains unsolved
- Editing a full podcast episode in Everycut; hiring an AI idea scout to mine Slack for angles
- Mapping all 1,469 Every essays in 3D and turning topic gaps into article pitches
- Three models taking turns improving one writing skill raised win rate from 18% to 100%; a solo-model control matched it
- Running Claude until it hit 46 GB of memory, prompting Dan to offer a spare 16 GB Mac mini
More from coding & agent
- Anthropic says no regression behind Claude Code usage complaints, blames heavy sessions — bcherny · 2026-10-29
- LlamaIndex launches Document Parsing Protocol: 140 rules to standardize OCR output for agents — llama_index · 2026-10-29
- exe.dev Ships WIF Support: Agents Make OAuth Token Exchange Finally Usable for All — davidcrawshaw · 2026-10-29
- Protecting action-taking agents from prompt injection: move defense out of the model layer — Informal-Dust4499 · 2026-10-29
- Designing agent approval screens: show the effect, not the narrative — ClickOk5811 · 2026-10-29
- "Babysit this for me" is now the number one request people make to AI agents — RichardsonDx · 2026-10-29