Category

AI and technology

Generative AI, LLMs, RAG and AI agents, from how they work — and where a person still has to check.

Learning path

Learn ai and technology from the ground up

Read from the top: each step makes the next one easier to follow.

  1. STEP 1How generative AI writesColumn
  2. STEP 2LLMsTerm
  3. STEP 3Tokens and contextTerm
  4. STEP 4RAG: letting AI read your documentsColumn
  5. STEP 5AI agentsColumn
  6. STEP 6MCPTerm
  7. STEP 7Using AI safely at workComing
Articles

13 articles

AI and technologyA shared way to opt out of AI training: what to decide for robots.txtOn October 30, 2026, an IETF draft proposed one robots.txt signal for telling AI companies not to train on your content, and Google said it aims to support it in 2027. What was settled, and what to decide for your own site.3 min readAI and technologyAttackers have started using AI: what Japan's recent breaches revealedIn October 2026, JPCERT/CC reported that attackers behind a wave of breaches in Japan had used AI to probe for known weaknesses and attack APIs. What has changed, and what a small business should review.3 min readAI and technologyThe Work People Still Need to Check in the Age of AIWhen you use AI answers at work, five kinds of content need a human check: facts and figures, anything sent outside, money, contract and HR decisions, personal information, and actions that cannot be undone. How to check without rubber-stamping, and how to assign who checks what.3 min readAI and technologyHow Much Admin and Accounting Work Can AI Take On? Sort It Task by TaskWhether AI can take on admin and accounting work depends on the task, not the job title. Drafting, reading and suggesting matches are easy to hand over; confirming amounts, making payments and legal judgments stay with people. A task-by-task table and what to decide first.3 min readAI and technologyWhen a Document You Give AI Becomes an Instruction: Prompt Injection BasicsPrompt injection is when text inside a web page or document you give an AI ends up acting as an instruction to it. A fictional example of how it works, why the impact depends on permissions, why RAG does not remove it, and what users and providers can check.3 min readAI and technologyAI Answers Missing the Mark? Four Things to Tell ItWhen a generative AI answer is not what you wanted, splitting your request into goal, context, material and answer format makes your conditions easier to convey. How to start short, fix what drifted, check facts, and what to confirm before pasting work documents.3 min readAI and technologyGiving Company Documents to AI? What to Check Beyond "Not Used for Training"Before entering company documents into a generative AI service, training is not the only thing to check. How to look separately at storage, access, connected services and deletion, what remains after removing names, and how to start with practice data.4 min readAI and technologyBefore Handing Work to an AI Agent: How to Decide Its PermissionsHaving AI draft a message and having it actually send or delete things call for different checks. How to break a job into operations, an example of initial permission settings, why instructions alone are not enough, and preparing logs and a way to stop.3 min readAI and technologyWhat Is RAG? Letting AI Answer from Your Own DocumentsRAG finds the parts of your internal documents that relate to a question and hands them to the AI along with it. How chunking, embeddings and retrieval work, why hallucinations drop, and why they do not disappear.5 min readAI and technologyWhat Is an AI Agent? How It Differs from a Chat AIA chat AI returns an answer. An AI agent uses tools to move a task forward. A step-by-step look at tool calling, APIs, MCP, how agents differ from workflows, what "autonomous" means, and where a person should stay in the loop.5 min readAI and technologyHow Generative AI Writes: Prediction, Not UnderstandingGenerative AI does not understand text the way people do. It predicts what comes next. A plain-language walk through tokens, large language models, training versus inference, the context window, and why confident mistakes happen.5 min readAI and technologyAI vs RPA: Fixed Tasks and Tasks That Need JudgementRPA repeats a set procedure exactly; AI reads and decides. A side-by-side comparison of RPA, workflows, AI and AI agents, how to separate fixed tasks from judgement calls, how to combine them, and where a person still needs to check.5 min readAI and technologyHow much of video editing can actually be automated?Silence removal, draft subtitles, scene detection, vertical crops and exports automate well. Choosing what stays, in what order, and where to leave a pause does not. Here is where the line sits in practice.5 min read