Agents

Your AI has amnesia, and the fix isn’t more memory — it’s better memory. Bài viết song ngữ Anh-Việt. Phiên bản tiếng Việt ở bên dưới. English Every time you have a long conversation with an AI assistant, something embarrassing happens behind the scenes. Around the 30-minute mark, the system quietly starts forgetting the beginning of your conversation. Not because it decided that stuff was unimportant — because it ran out of room. These systems don’t have memory. They have a sliding window, a fixed-size buffer that moves forward as the conversation grows, dropping older context off the back end. That brilliant setup you laid out in the first five minutes? Gone. ...

The same task, two radically different approaches. One says “write code to orchestrate the LLM.” The other says “write markdown to teach it.” Both produce agents that reason, use tools, and complete complex work. Knowing when to reach for which is the skill that matters in 2026. The SDK Way: Agents as Code Agent SDKs — OpenAI Agents SDK, Claude Agent SDK, LangGraph, CrewAI — let you build agents programmatically. You register tools as functions, write instructions, and the SDK runs the loop: prompt the LLM, execute tool calls, feed results back, repeat. ...

Part 2 of Two Roads to AI Agents. This time we apply the framework to something real. In Part 1, we drew the line between Agent SDKs (code orchestration) and Agent Skills (markdown knowledge). Now let’s see where that line falls in practice. We’ll walk through a pipeline that turns any article into a narrated MP4 — and at each step, I’ll label whether it belongs in SDK code or Skill knowledge, and why. ...

LLMs Don’t Actually “Call” Tools Here is a fact that surprises most people: LLMs cannot run code. They cannot call APIs. They cannot read files. All they do is predict the next token — they output raw text, one piece at a time. So how do they “use tools”? The trick: structured wishes During training, models are fine-tuned on augmented datasets that include tool call examples. The training pipeline works roughly like this: ...