The best microbooks of the week, selected by our team. Stay tuned to the latest trends in the world of literature and knowledge.

Radar 12min

Radar 12min is a narrative audio journalism show produced by 12min. Each episode takes one relevant topic and turns it into a dense, direct story — no filler, no fluff. You get verified facts, real context, and a closing section that helps you think through what to actually do with the information. This episode is about artificial intelligence... but not the kind you've already heard about a hundred times. It's about what AI is doing right now, in the real world.

Five real stories about what AI is already capable of doing

Radar twelve min tracks what is shifting in the worlds of technology, science, and business and translates it into useful reading. No alarmism, no hype, just enough context for you to form your own opinion about what is coming.

TRIBE V2 — THE MACHINE THAT LEARNED TO READ THE BRAIN

New products, new technologies. See the future with the eyes of ingenuity.

Technology & Innovation

Salman Khan reveals how generative AI can finally deliver what education has always needed: a world-class personal tutor for every student on the planet. Drawing on his work with OpenAI and the creation of Khanmigo, Khan shows how AI tutors use the Socratic method to guide reasoning instead of handing out answers, how they unlock creativity across the humanities, and why the "centaur" professional — half human intuition, half machine power — is the blueprint for the future workforce. A practical, optimistic roadmap for turning AI into education's greatest ally.

Brave New Words

A practical guide to the full lifecycle of machine learning systems — from aligning business and model metrics to shadow deployment, data drift monitoring, and continuous learning in production.

Designing machine learning systems

Meta invested fourteen billion dollars, built a superintelligence lab from scratch, and launched Muse Spark, its new AI model. This Radar breaks down what it does, where it beats and where it loses to OpenAI, Google, and Anthropic, why Meta abandoned open source, and what changes for anyone using WhatsApp and Instagram.

Muse Spark... Meta spent 14 billion on its new super AI

This Radar investigates Kuse AI and its product Junior, a virtual employee that operates twenty-four hours a day inside the tools companies already use. The text analyzes how the technology works in practice, what infrastructure supports it, who is already using it, how much it costs and what economic data says about the real impact of automation on the job market.

Virtual employee: two thousand companies are already in line

Stuart Russell warns that the real danger of superintelligent machines is not malice but blind obedience to badly defined goals. This microbook lays out three principles for building Artificial Intelligence that stays humble, uncertain, and permanently aligned with human values.

AI For Our Favor

While five hundred million people were chasing Pikachu, the company behind Pokémon GO was quietly amassing thirty billion images of the real world. Now, that database powers delivery robots that navigate sidewalks with centimeter-level precision. This Radar investigates how a mobile game became the largest visual crowdsourcing project in history, who owns that data today, and why it matters to anyone who has ever downloaded a free app.

Pokémon Go, the CIA, and thirty billion images: The invisible price of a free game

An Austrian programmer built in one hour the software that is transforming China. OpenClaw is an open-source AI agent that does not just talk... it acts: it replies to emails, operates programs, buys tickets, and works around the clock. Millions of Chinese are "raising lobsters," local governments are offering multimillion-yuan subsidies, and a new model of one-person companies is emerging. But the security risks are real. This Radar covers both sides of the revolution.

AI (Artificial Intelligence)

Most Read

Newest