This week's useful read is the split between teams celebrating visible AI usage and operators asking whether the spend produces reviewed, measurable work.
What mattered this week
Augment Prism routes coding turns for cost and quality
Official Prism launch note on per-turn model routing for coding work, framed around cost control without forcing teams onto one model family.
Takeaway: Prism is tokenmaxxing discipline in product form: route expensive coding turns only when the expected quality gain justifies the extra cost.
Read source noteAnthropic raises Claude Code limits with new compute
Anthropic ties higher Claude Code and API limits to new compute capacity, making capacity itself part of the agent-product story.
Takeaway: More compute raises the ceiling for token-heavy coding workflows, which makes spend controls and outcome checks more important, not less.
Read source noteHR experts warn token dashboards are weak productivity metrics
Canadian workplace experts argue token dashboards can show AI adoption, but they are weak measures of output quality or business value.
Takeaway: It reinforces the core warning: high token use can prove experimentation while still saying little about whether the work improved.
Read source note
