Reflections & stories
Ideas, experiences and learnings about education, technology and coexistence.
Showing 1585–1596 of 1,735 articles
Security begins when reporting an error is possible when an application asks for information: Privacy and data
An error report can reveal sensitive data while being essential to correcting an app. A safe channel should request the minimum, explain who reviews the case, and allow information to be corrected or withdrawn.
Read article
What works for most people can exclude someone when an application asks for information: Privacy and data
An option that seems simple for most people can exclude someone who shares a device, needs another language, or cannot provide every datum. Apps need real alternatives, limits, and review.
Read article
A rumour travels quickly when it tries to care when an application asks for information: Privacy and data
A rumour can seem like a useful alert and still turn a query into a file about a person. Applications must distinguish necessary information from surveillance, provide verification, and protect context.
Read article
Conflict is not resolved by adding another screen when an application asks for information: Privacy and data
A conflict is not improved by turning every doubt into another screen and another collection of data. A responsible app limits evidence, offers human review, and protects the context of those involved.
Read article
An exception reveals a system's quality when an app asks for information: Privacy and data
Exceptions show whether an app can protect data when the ordinary path no longer works. A good system limits access, explains the decision, and enables repair without turning a singular case into surveillance.
Read article
A useful number can still lead us to the wrong place when an app asks for information: Privacy and data
Metrics help detect failures, but they do not justify collecting every available datum. A responsible app measures what is necessary, explains its limits, and retains human alternatives.
Read article
Accepting does not always mean being able to choose when an app asks for information: Privacy and data
Acceptance is meaningful only when an app explains which data it needs, why it will use them, and provides real alternatives. Privacy needs understandable decisions, not default permissions.
Read article
The first question must come before the tool when an app asks for information: Privacy and data
Before asking for data, an app should explain its purpose, offer an alternative, and limit collection to what is essential. The first question protects privacy and improves the service.
Read article
Participation changes the question and improves the answer around an automatic response: Everyday artificial intelligence
Participation turns an automatic response into a service that can be corrected: it makes room to identify errors, add context, and decide what improves without shifting every burden to the recipient.
Read article
Care also means not monitoring everything around an automatic response: Everyday artificial intelligence
Care around an automatic response does not mean recording every action: it means limiting data, providing a human route, and reviewing only what is needed to correct harm.
Read article
Supervision requires the power to stop and correct an automatic response: Everyday artificial intelligence
Supervising AI is not watching a screen: it requires authority to stop a response, correct its effects, and explain the change to affected people.
Read article
After launch, the important work begins — before an automatic response: Everyday artificial intelligence
Launching an AI tool does not close a project: it starts error observation, support for affected people, and verifiable improvement of every automatic response.
Read article