Notes on everyday assistants: writing, tone, and the tradeoffs labs make when they chase agents.
SynthID-Text lives in synonym ties. Most Claude flagships on the board still skip it. Gemini text usually carries it.
Turning on deeper reasoning buys helpfulness and quietly spends empathy. Sibling DeepSeek modes and the board’s helpfulness–empathy gap show a second tax next to the writing one.
Labs optimized internal reasoning for token-efficient agents. That same compression bleeds into final prose, so creative writing got worse even as helpfulness climbed. GPT-4.5 is still the writing high-water mark.
The same chat interface that runs at chatio.dev/chat is now on GitHub: local-first, bring-your-own-keys, and self-hostable under Apache 2.0.
9 min
6 min
8 min
4 min