Thriving
Health improving, rights advancing, lives getting better.
What bias does this lens correct?
Bad news feels more important than good news. That's the negativity bias. Researchers like Rozin & Royzman (2001) and Baumeister et al. (2001) showed that negative information receives disproportionately more attention and weight. News media amplify this: when the world is described only through crises, progress becomes invisible.
What does this lens find?
Articles showing people doing better: improvements in health, quality of life, wellbeing, education, or social connection. Not feel-good stories, but documented progress with measurable results.
What you will and will not see here
You will see
- Something delivered and received, so that people you can identify are better off now, in health, community, justice or rights, in a way that looks like it will last.
- One person clearly better off is enough.
- A recent, observed improvement for a named group of people in a named place, such as a falling rate of illness, including research that documents it.
- A newly running facility. A public service or a cultural milestone only when the article shows people newly better off.
- Good news from a hard place.
- Evidence that people actually got it: figures, people saying what changed, a reporter who saw it, or a record outsiders can check.
You will not see
- Stories whose news is a harm, such as a crime, a death or a worsening number, even if a hopeful programme appears later in the story.
- A response to a harm that has not yet repaired anything.
- Money pledged, or a facility announced but not yet running.
- A law, award, ruling or agreement that has not yet reached people.
- Corporate success, including record profits.
- An organisation's own press release or statement, a government's included, as the only evidence about the people it helped.
- Predicted benefits.
An article can fit more than one lens. We publish it in the narrowest lens it passes. The lenses look for evidence that something happened. How well a story is told plays no part, except for a small share of the Discovery score.
Scoring dimensions
Each lens evaluates articles on six dimensions. Together they form the profile you see in the radar chart.
Improvement in health, quality of life, or personal welfare
Strengthening of community bonds and solidarity
Progress in fairness, equality, or human rights
Reliability of the claims based on available evidence
How broadly and fairly benefits are distributed
Likelihood the improvement will last, not fade
What this lens actually picked
Three stories the Thriving lens selected, spanning its range. The last one barely cleared the bar — it is here on purpose. A method is only judgeable if you can see the cases it nearly rejected. Headlines and summaries are written by AI; the scores are what the lens produced.
Notice that a story can pass on one strong dimension while scoring poorly on another. The lens does not require a story to be good in every way — it requires enough of what this particular lens is looking for. That is also how it gets things wrong, and why every story links back to the publication that reported it.
How does scoring work?
Our AI analysis system evaluates each article on the dimensions above with a score from 0 to 10. The weighted average determines whether an article passes the lens. Articles below the threshold are not shown. Not because they are bad, but because they do not fit strongly enough what this lens looks for.
Limitations
These dimensions are designed criteria informed by existing research. They are not established psychometric scales. The AI model can make mistakes: missing relevant articles or letting irrelevant ones through. The scores are a selection tool, not a definitive judgment of a story's value.
Want to know more?
Read how we work for the full picture, or browse the source code on GitHub.