This blog explores how we could build and organise things better — from AI and safe automation to everyday design and society. The articles develop ideas, proposals and the reasoning behind them.

Suomeksi →

Start here

The personal comedy model

What happens when everyone has a personal AI comedian that learns exactly what makes them laugh? Novelty, a value estimate and effects on shared humour.

Read more →

Latest articles

  • A robot ferry could turn water into a shortcut

    The opposite shore is visible from your window, but driving there requires a long detour. One possibility would be a network of small electric robot ferries carrying groceries, parcels, and tools over short water crossings. The vessels would not need to operate in every weather or throughout the year. They would provide inexpensive shortcuts when conditions were good enough and remain docked otherwise. Read more →

  • A prescription for your hair – take a good haircut to your next barber

    A good barber gradually learns what works for your particular hair. When you change barbers, that knowledge usually does not come with you, and the explaining starts again. One possibility would be a haircut prescription that travels with the customer: a map recording the instructions for a successful cut and the particular features of their head. You could show it to a new barber much as you would take an eyeglass prescription to a new optician. Read more →

  • A peace machine – revealing concessions at the same time

    In an argument, both people may be willing to make concessions, but neither wants to go first. One possibility would be an AI assistant that would help each person clarify their position and prepare a concession privately. Once both had approved their own messages, they would be revealed at the same time. The aim would be to make a joint conversation easier to start and find solutions without either person first having to feel they had lost. Read more →

  • A browser could stop accidental taps on jumping pages

    You are about to tap a link when an ad appears and the page jumps. Your finger lands where the link was a moment ago, but the browser opens something else. I propose a browser safeguard that recognizes this situation and discards the suspicious tap. The user lifts their finger and tries again once the content has settled. Read more →

  • An appeals CAPTCHA: appealing stays free, but not a hundred times over

    Websites often give you a small task, such as spotting the traffic lights in a set of pictures. For a person it takes a second, but for a machine trying to log in thousands of times it is a brake. I propose the same idea for appealing against decisions made by public authorities. When someone files an unusually large number of appeals, each one would have to be confirmed in person with an official, on site or over a video call, explaining in their own words which decision they are appealing and why. For one genuine appeal that is a... Read more →

  • A general offence of spiteful harassment: a draft for the Finnish Criminal Code

    The law today punishes stalking, threats and spreading lies, but stalking requires that the same person is harassed repeatedly. What slips through is harassment where each victim gets their share only once. Picture a video maker who goes out to embarrass, startle or provoke strangers on camera, hundreds or thousands of times, and earns ad money from the videos. None of the victims has been stalked, yet harassment is the whole business model. I propose a new, general offence of spiteful harassment: it would be a crime to act in a way whose clear main purpose is to torment someone... Read more →

  • The early adopter investment paradox: the service may disappear, the record should not

    A new app often becomes useful only once it has gathered lots of data and lots of users. The first users do that work: they enter their information, learn how the thing works and trust that the effort will pay off one day. Yet they are exactly the ones who lose the most if the app is shut down, because their work stays locked inside a service that no longer exists. I propose turning this around: the earlier someone joins, the less they should have to invest, and the data they enter should be theirs, in an open format that... Read more →

  • The wall bed gap should be closed by design, not just by rules

    In many Finnish daycare centres, children take their naps in beds that are folded up against the wall at the end of the day, a bit like the door of a cupboard. In spring 2026 in Vantaa, nobody noticed that one child was still asleep in a bed as it was being raised. The child slid into the gap between the bed and the wall and died. So far the response has been twofold: staff are instructed not to raise the beds while children are present, and the most dangerous models are being replaced. I propose a third option: close... Read more →

  • The personal comedy model

    AI can already write jokes that people find genuinely funny. So far the question has mostly been whether a machine can be funny. I ask a different one: what happens when everyone has their own joke machine that learns from every chuckle and every skip what makes them laugh, and immediately invents the next joke just for them? Video apps already search existing content for what you are likely to enjoy. This would make the content from scratch for you, a bit like a comedian whose entire audience is you. That is why the question is not only about entertainment... Read more →

Subscribe via RSS