Vol. 1No. 55
Updated 1d ago
Agents won't arrive as robot coworkers. They'll arrive as a thousand small handoffs, and the people who learn to delegate, inspect and price that work will set the terms for everyone else.
by Dev Raman
An old classroom trick for modern models: before any big task, ask the AI to restate the job, flag its risky assumptions and ask you three questions.
by Jonah Kim
Invisible watermarks on AI content are easy to strip and easy to skip. Documenting where authentic content came from is a quieter, more promising idea.
by Dev Raman
by Priya Sethi
Long conversations with the people building with AI — founders, artists, researchers and rule-makers.
Hosted by Theo BrooksUses versus models, rules that expire, sandboxes that aren't loopholes: Theo Brooks talks with a nonpartisan researcher who tries to get both sides of the AI debate in one room.
The illustrator and composer draws every album page by hand and builds her ambient tracks with a little help from generative tools. A talk about taste, credit and why 'unfinished' is a feature.
One prompt, one idea, every weekday. A column by Jonah Kim.
One explanation rarely sticks. An analogy, a mechanism and a worked example, plus a test question, usually does. A former teacher's favorite learning prompt.
by Jonah Kim · 3 min read
Imagine it's six months from now and the plan has flopped. This prompt turns that grim fantasy into ranked risks, early warning signs and one cheap test to run this week.
by Jonah Kim
Agents, workflows and the quiet automation of everything.
A 95%-reliable step sounds great until you chain twenty of them together. A quick lesson in compounding error, and how builders fight back.
You don't need a brand name; you need the right kind of tool. We built the same three agents with each of the five main kinds of agent builder and scored what actually mattered.
Prompt of the Day· Nov 9
An analogy, a mechanism and a worked example, plus a test question, for learning anything hard.
Fill in the 3 highlighted variables before you run it.
I'm trying to understand {{concept}}. My background: {{what I already know}}. Explain it three ways: 1. An analogy from everyday life ({{an area I know well, e.g. cooking, football}}). Then say where the analogy breaks down. 2. The mechanism: how it actually works, step by step, in plain language. 3. A worked example with real numbers or a concrete case. Then give me one test question that checks whether I really understood it, not just memorised it. Don't give the answer until I try.
Ask Ideabro · AI
Every answer cites the stories it came from. Answers are drawn from 40 stories in our archive.
AI answers can be wrong. Check the cited stories before you act on anything important.
AI-native companies, business models and ideas worth stealing.
Seats, credits, tasks or outcomes? Every AI pricing model makes a promise about who carries the cost of the machine. Here's how to choose yours.
A meeting-notes startup found fierce loyalty on construction sites, then left that niche to chase everyone. A post-mortem on bundling, wedges and the danger of going generic.
Films have end credits, albums have liner notes, books have acknowledgments. Generative tools are pushing every creative field to rethink what 'made by' means, and some artists are writing new rules.
From the story
Priya Sethi
Tools we actually use, rated by the newsroom.
A research assistant that greys out every sentence it can't source.
A pull-request reviewer that explains every suggestion and points at the exact line it's worried about.
A spreadsheet assistant that explains a formula in plain English before applying it.
Removes room echo, hum and keyboard clatter from spoken recordings.
A writing partner that argues with your outline before it writes a word.
A visual canvas for building agents, with approval steps built in from the start.
The Ideabro Forum is where readers trade ideas, prompts and builds. Be generous, be specific, be kind.
Join the conversationIdeas from readers and the Idea Lab. Upvote the ones you'd build.
Idea #2
A ready-to-run kit for libraries and community centres to host a weekly 'is this AI answer true?' help table.
Grace Whitfield81upvotesIdea #1
A rota builder for tiny shops that treats staff preferences as hard rules and shows every rule it checked before you publish the week.
Ruth Abernathy64upvotesIdea #11
Helps early-career developers turn AI-assisted projects into a portfolio that shows judgement, not just output.
Lucía Romero59upvotesIdea #3
A shared, teacher-written set of science questions for testing whether an AI tutor teaches or just gives the answer.
Hannah Lindqvist57upvotesIdea #7
Point it at a woodworking video and get a materials list and cut list, with timestamps for every measurement.
Oliver Grant52upvotesIdea #4
Upload a shipment's documents and get a checklist of mismatches between invoice, packing list and declaration before customs does.
Kwame Asante49upvotesGot a hunch? Let the Lab turn it into a pitch in seconds.
Generate your own ideaBuilding something with AI? Tell us. The best pitches become stories.
