Guozhen AIGlobal AI field notes and model intelligence

Realtime AI News

Stop counting tokens, start counting what ships

Stop counting tokens, start counting what ships. When a developer gives an AI agent a coding task, the agent does more than answer once. It plans, generates code, runs tests, and repeats the process to correct errors. A single task can use 10 to 50 times more tokens than a basic chat prompt. The more useful...

Published
AI动态:Stop counting tokens, start counting what ships
Image source: technologyreview.com

According to technologyreview.com, Stop counting tokens, start counting what ships.

When a developer gives an AI agent a coding task, the agent does more than answer once. It plans, generates code, runs tests, and repeats the process to correct errors. A single task can use 10 to 50 times more tokens than a basic chat prompt. The more useful...

The signal matters because AI capabilities are moving into more specific product, infrastructure, or business workflows.

The next things to watch are availability, pricing or access limits, and whether the update creates a measurable workflow change for builders or enterprise users.

Why it matters

This update reflects the continued movement of AI capabilities into concrete product, platform, and industry contexts.

AI NewsAgent
Back to realtime news

Nearby Updates

All

11/02, 21:36

SenseTime launches SenseNova 6.8 Flash, a multimodal agent model that is free for now

SenseTime has launched SenseNova 6.8 Flash, a multimodal agent model that is free for a limited time and built around delegated intelligence: users set a goal and the model completes the workflow for sign-off. SenseTime says it ranks first on TeamBench and DataAnalysis2.0 and second on several other office and browsing benchmarks.

11/02, 21:57

OpenAI moves to merge ChatGPT's Chat and Work tabs

OpenAI is preparing to merge the Chat and Work tabs inside ChatGPT, starting with a select group of users for feedback before a wider rollout. An email to users also says ChatGPT and Codex will then draw usage from the same allowance pool, so some people may need to wait for resets more often.

11/02, 22:00

Aion Launches an AI Agent for Business Finance Across Banking, Payments, Cards, and Credit

Aion has launched an AI agent aimed at business finance work spanning banking, payments, cards, and credit operations. The pitch moves agents out of coding demos and into back-office functions where the work is repetitive, spread across systems, and auditable.

11/02, 22:05

Google Shuts Down Jules, the Coding Agent That Promised to Take On the Work You Didn't Want

TechRadar reports that Google has killed off Jules, its AI coding agent marketed as a way to hand off development chores developers would rather not do. The shutdown lands while coding agents are the most contested category in AI tooling, making it a signal about how vendors are pruning their own agent lineups.