General AI
MIT uses AI to map border deaths from police reports
MIT Technology Review AIhigh confidence
Why it matters
- →Demonstrates a scalable workflow for extracting structured geospatial data from unstructured police reports using LLMs.
- →Shows how AI can reduce the cost and time of investigative journalism involving large volumes of public records.
- →Highlights the practical application of API-based model access for specialized data extraction tasks in non-tech industries.
MIT Technology Review and the Times of San Diego have published the first comprehensive map of migrant deaths near US border surveillance towers, leveraging AI to process fragmented government records. The team used Anthropic’s Claude API via API to extract location coordinates from over 4,000 pages of Texas sheriff reports, a dataset previously inaccessible due to its unstructured nature. After manual verification and filtering, the analysis included more than 1,500 confirmed cases dating back to 2015. The project highlights how generative AI can bridge data gaps in public interest journalism by automating the extraction of key fields from legacy paper-based or PDF records.
Read originalThe story around this
Earlier coverage that leads up to this article, and what followed. Lines connect each piece to the closest one after it, converging here.
