Catching up on the news with a Dcipher Research Bot

A Dcipher Research Bot is a life saver whenever you have questions about what is reported and discussed in news or social media. In this blog post, we train a bot that can tell us about the news reporting from the U.S. primary elections.

While some other popular AI chatbots lack knowledge of everything that has happened in the world during the last several years, a Dcipher Research Bot will always be up-to-date with any material you feed it, which can be news articles as recent as from the last hour.

Let’s say we want to catch up on what’s been happening so far during the primary elections in the United States. There are hundreds of thousands of news articles on the topic out there waiting to be read, but we may prefer to let a bot go through them rather than doing it ourselves.

How do we go ahead with this? Easy: we just log into Dcipher Analytics and click “Create a new Research Bot” from the Research Bots view. This brings up the set-up wizard, which let’s us define what it is we want our bot to read for us. We choose to train our bot on news, and then specify that we are interested in news articles written in English, mentioning either “primary elections” or “primaries”, and published in the United States on or after January 15 this year (the date of the Republican caucus in Iowa, with which the elections season started):

After about 15 minutes’ wait, our Research Bot is ready. Let’s start by asking the bots who the current presidential candidates are (as of February, 2024, when this blog post is written):

If we have not been following the news since the primaries started, we may wonder what has happened to two other Republican candidate whom we had previously heard about:

There are of course an endless number of fact questions that you might want to let the bot answer based on the news articles it has read. Why not train your own Dcipher Research Bot and try for yourself?

We may also ask the bot questions that are not solely about what events have taken place. Has the bot read anything surprising, for example? While our Research Bots may not be able to experience the feeling of surprise themselves, they can pick up fenomena that have been described as surprising in the source texts. When asking our bot if there were any surprises in Iowa, the first state out in the Republicans’ calendar, its reply does not refer to the election outcomes—indicating that the results have not been described as surprising in the news reports—but instead to the extremely skewed racial composition among Republican voters and to the extremely cold weather.

Something else we can ask about regards sentiments towards the leading candidates, which the bot can tell is mixed within both parties:

We have obviously just been scratching the surface with these questions. Given the vast number of articles that the bot has read, we could dig much much deeper into the details from the political reporting—but that we leave to you to train your own bot to find out.

By the way, Dcipher Research Bots can handle news articles and other texts in a large number of languages, while answering your questions about them in English. Why not train one to help you catch up on the news reporting ahead of this year’s presidential elections in Iceland, Mauretania, or North Macedonia?