Practical AI for solving real problems and creating value
The article reflects on an Alibaba Cloud AI productivity roundtable, summarizing speakers' insights on how AI can expand personal capability, automate expert workflows, and create lasting competitive advantages that depend on integrating industry knowledge, judgment, and execution rather than merely using tools.
The article identifies five groups—process‑followers, slow learners, tool‑resisters, uncritical thinkers, and poor communicators—who risk being outpaced by AI, and outlines the traits of a "super individual" who can thrive by continuously learning, thinking critically, and leveraging AI tools effectively.
The article explains how the author packaged repetitive photo‑album download tasks into two Skills—alltuu-downloader and photoplus-downloader—so that an AI Agent can automatically fetch original images from cloud albums, saving users from clicking each picture individually.
The article examines why a model’s advertised context window (e.g., 128 K or 1 M tokens) does not guarantee effective long‑context reasoning, summarizing the RULER framework that breaks long‑context ability into retrieval, interference resistance, multi‑hop tracking, aggregation, and multi‑answer recall, and offering practical guidance for evaluating and using such models.
This article walks through enabling Codex’s new desktop‑pet feature, selecting a built‑in pet, installing the Hatch Pet Skill from GitHub, creating a custom pet by describing or uploading an image, and using a generic cyber‑pet tool for other AI coding assistants.
The article recommends five versatile AI Skills—find-skills, skill-creator, skill-vetter, dokobot, and superpowers—explaining their use cases, installation commands, and how they help automate workflows, while encouraging readers to create custom Skills when none match their needs.
claude-trace intercepts Claude Code’s Anthropic API calls, logs them to .jsonl, and generates a self‑contained HTML report that displays system prompts, tool definitions, tool outputs, and full request/response details, with options to include all requests or index the logs for searchable summaries.
The author explains that most AI PPT generators fail to deliver consistent visual style, reliable Chinese text, and easy editing, and shows how GPT‑Image 2 combined with a custom‑defined visual style and a reusable Skill can reliably produce high‑quality, style‑consistent slides without complex prompts.
The article outlines thirteen actionable strategies—ranging from choosing the right billing plan and trimming context to using layered models, caching, and proper output prompts—to dramatically reduce token consumption and overall expenses when working with AI services.
The article compares DeepSeek V4 Pro and Flash, explains how to select the right model based on capability versus cost, cautions against relying on flashy demos, praises the restrained release, and provides step‑by‑step instructions for API integration and tool configuration.
