@nyt-first-said is a twitter bot which tracks when the New York Times publishes a word for the first time in its history.
It looks like this:
subtweeted
— New New York Times (@nyt_first_said) January 30, 2018
Some points of inspiration for the project are Allison Parrish's @everyword bot, and @nyt_diff, a twitter bot inspired by the NewsDiffs project.
bot idea: whenever the nyt publishes a lowercase word for the first time in it's history, tweet it.
— Max (@maxbittker) March 1, 2017
An important thing to understand about @nyt-first-said is that the process to extract these words is a best effort undertaking, and not neccesarily 100.00% accurate or exaustive.
For instance, the bot takes some steps to throw away less-interesting first words – such as words containing capital letters (proper nouns) or certain punctuation (URLs). Despite this, it still picks up a lot of typos and nonsense... but I like to think of that as part of its charm.
There are a few sibling bots, including @nyt-said-where, which replies to each tweet with a paragraph of context from the source-text and a link to the article.
You can find a more detailed explanation about how this bot works on the
github repository -
and that's also the best place to file bugs if something is wrong.
Thank you for reading and feel free to let me know @maxbittker if you think there are other important questions I should answer on this page!
If you're curious, I've done some other projects, and you can find some of those on my website.