
Anthropic launches Claude Sonnet 5 as a cheaper way to run agents
THE SO WHAT
Sonnet 5 framed explicitly as a cheaper agent engine—versus Opus, GPT-5.5, and Gemini Pro—turns agent cost per decision into a competitive metric. If you’re running multi-step or multi-agent systems, re-run cost-per-workflow comparisons this week; model choice is now an operating expense lever, not just a quality call.
READ THE SOURCE
MORE FROM THE WIRE
Applied AIGoogle’s NotebookLM can sum up your research in a TikTok-style clip
Research workflows are being rebuilt around short-form, auto-produced explainers — 60‑second vertical clips as a first pass over dense source material. Teams that rely on internal docs and knowledge bases should assume video-native consumption and design their knowledge architecture for remixing, not just reading.
Applied AIGrindr Is the Latest Dating App to Hook Up With AI
Dating apps adopting AI move recommendation and conversation from UX tweaks to core product logic — whoever controls the matching and messaging models controls the business. If you run a marketplace or social graph, assume users will expect AI-native onboarding, safety, and communication features within the same app, not bolted on via third parties.
Applied AINewsom strikes Anthropic deal to get California government half price Claude AI access
A 50% discount for statewide Claude access turns AI from pilot spend into line-item infrastructure for California agencies. If you sell into government or large enterprises, expect procurement to start negotiating AI access like cloud contracts — volume deals, standardized models, and pressure on per-seat and per-token pricing.
Applied AIGoogle introduces a faster, cheaper image generator with Nano Banana 2 Lite
Cheaper, faster image generation pushes AI art from creative experiment to default asset pipeline for marketing and product teams. If your business depends on stock imagery, basic design work, or UGC filters, assume margin compression and differentiate on brand, workflow, or rights — not raw generation.