Iβve spent a lot of time investigating shitty AI cynically used to pursue vulnerable people in welfare systems. When I talk about this work, I invariably get asked: is it possible to make these systems fair?
11.06.2025 19:00 β
π 10
π 7
π¬ 1
π 0
Our investigation epitomizes the storytelling possibilities when passionate journalists from resource-strapped newsrooms and regional data geeks join forces. We hope the story leads an overhaul of NGP but also sparks wider scrutiny of sustainable forestry programs across the region and the world.
23.05.2025 02:51 β
π 4
π 0
π¬ 0
π 0
European Economic Labour Force Survey, Immigrant Brain Waste Data
The dataset is based on microdata from the European Labor Force Survey (ELF) for all EU countries except Germany, plus Norway, Switzerland, Iceland and the United Kingdom (until 2019). The survey c...
5. Our impact strategy recognizes that journalism is one part of helping to drive systemic change. Our strategy seeks to connect with policymakers, academics and CSOs to build a coalition. Access our data here purl.stanford.edu/vv199dn8443 thanks to @biglocalnews.bsky.social
24.12.2024 04:50 β
π 0
π 0
π¬ 0
π 0
4. Our evolving editorial objectives and impact goals were in constant conversation as we conducted the investigation and enabled us to stay focused on why we were doing the investigation in the first place and maximize our potential contribution to solutions.
24.12.2024 04:50 β
π 0
π 0
π¬ 1
π 0
3. Systematic failures are gnarly and nuanced. Publico is a collaborative partner that was aligned with our impact goals. We agreed to multimedia content that brings readers into the lives of migrants and accessible data explainers that help audiences understand the scale of the issue.
24.12.2024 04:50 β
π 0
π 0
π¬ 1
π 0
2. We wanted audiences to realize the economic potential of more inclusive approaches to migration. A statistical model estimates the economic cost to show brain waste harms professional migrants themselves and national economies that fail at integrating these professionals into the labor force.
24.12.2024 04:50 β
π 0
π 0
π¬ 1
π 0
1. We wanted to find a mainstream, centrist audience that would not otherwise know about potential value of migrants to society. France 3, the third most watched television station, got us outside of our media bubble. We invested in relationships to create new alliances outside of the echo chamber.
24.12.2024 04:50 β
π 1
π 0
π¬ 1
π 0
After each part of the series, Iβve shared lessons learned. For Part 1, I focused on how we built and refined our data pipeline. In part 2 I explained how we prioritized the people over the numbers. For Part 3, I a, sharing how our impact strategy shaped our investigative approach.
24.12.2024 04:50 β
π 0
π 0
π¬ 1
π 0
Part 3 of the @lhreports.bsky.social investigation in collaboration with @elpais.com @unbiasthenews.bsky.social and @financialtimes.com , now with partners from @france3fr.bsky.social and @publico.es is out!
24.12.2024 04:50 β
π 4
π 0
π¬ 1
π 0
Amazing work by @gabrielgeiger.bsky.social on the latest @lhreports.bsky.social suspicion machines investigation. This is a crazy tale of AI ethics washing: The welfare agency had been told that it's tool was crappy. It responded by making up a ludicrous fairness definition.
27.11.2024 18:36 β
π 20
π 8
π¬ 1
π 0
Hello Bluesky friends! We'll be posting here more often going forward
In the meantime, sharing a starter pack of current/former staff and contributors to Lighthouse Reports investigations. Will update as we go!
go.bsky.app/J6a68Yg
13.11.2024 17:29 β
π 74
π 30
π¬ 0
π 4