"For all the anxiety over how AI will upend journalism as we know it, I continue to believe its best immediate uses in our profession remain exceedingly mundane. Among them, extracting structured data from unstructured text: identifying people, places, and events buried in articles; applying metadata and complex taxonomies that reporters and editors don't have the patience to maintain; normalizing things like locations, spellings, names, and entities so they line up cleanly with external databases."
"It's tedious and unglamorous work. It also happens to be the foundation of some of the most impactful product, storytelling, and business model innovations our industry has launched over the last 20 years. The New York Times turned archived recipe articles into a product that now forms a core pillar of its bundle strategy. Politifact won the Pulitzer Prize and completely rebooted the concept of the fact check. On the business side, the Washington Post and others use taxonomies to target ads contextually"
AI's most practical near-term newsroom applications involve extracting structured data from unstructured text, identifying people, places, and events, applying metadata, and managing complex taxonomies. Normalizing locations, spellings, names, and entities enables clean alignment with external databases and reduces manual editorial maintenance. Classification and normalization work underpins product, storytelling, and business-model innovations such as turning archived recipes into revenue-driving products and transforming fact-checking into award-winning reporting. Taxonomies also enable contextual ad targeting and brand safety. Large language models and supporting technologies have lowered technical and cost barriers, making structured-data extraction achievable for many news organizations.
Read at Nieman Lab
Unable to calculate read time
Collection
[
|
...
]