How The New York Times Reviewed 7 Million Words Uttered by Trump
Two Times journalists, blending A.I. analysis and traditional reporting, evaluated public speeches to spot trends.
TL;DR
- An analysis combined AI and traditional reporting to study President Trump's public speeches.
- The study found that Trump's speeches have become more profane, absolutist, longer-winded, and less coherent over the past decade.
- Journalists used AI to identify potential trends and then verified the findings through traditional reporting methods.
- The methodology involved translating queries into computer code and prompts to pull excerpts, which were then reviewed in context.
- Basic statistical analyses of word frequencies were also used to assess vocabulary range.