Gazing towards the future
What does today’s world believe will happen in year 2030? In this blog post we use Dcipher’s Insight Booster Toolkit for a foresight exercise to map out what news media around the world are talking about when they mention that year.
Dcipher’s content landscaping workflows are unbeatable when it comes to quickly getting an overview of large volumes of unstructured texts. In this example, we will collect text passages in English-language news articles from the last 12 months that mention the phrase “in 2030.”
What text passages containing “in 2030” tend to have in common is that they cover some aspect of what someone expects the world to look like that year. As 2030 is often used as a milestone when people think about and plan for future developments, hundreds of thousands of news articles referring to 2030 are published each year. They usually focus on just one particular phenomenon, but if we go through a large sample of articles and then structure our observations, we will be able to make a map of what today’s world thinks about the world in 2030.
Doing that manually would a very tedious process. Using Dcipher, on the other hand, we just need to run a “Content landscaping of news” workflow and follow the on-screen instructions to assign an AI the task of creating a visual overview for us. After about 15 minutes we receive an email notifying us that we have an interactive report waiting for us. This is what it shows after processing a random sample of 10,000 articles:
Click here to access the interactive report.
We are looking at what we at Dcipher Analytics call a “content landscape.” It is an automatically generated visual representation of the processed text content, which has been structured thematically. You can think of the content landscape as topographic map, on which hills represent thematic areas in which there are many articles with similar content. The circles represent topics (circle sizes corresponding to topic sizes), with related topics sharing the same color. To help us immediately get a sense of what we can find in different parts of the landscape, a set of AI-generated labels describe the content at a very high level.
From the labels, we can see that the landscape contains areas from climate action, via sport events, to space exploration. Several labels are related to sustainability in one way or another, which is not strange given the prominent position that the year we are looking at has in the United Nations’ 2030 Agenda for Sustainable Development.
The interesting part comes when we start to look at the identified topics more closely. Hovering over the map’s colored circles with the mouse pointer, we see a description of what the topic it represents is about. (The “Global efforts…” topic seen in the illustration above is an example.) Wherever we want to learn more about a topic, we can click on its circle to get an AI-generated summary together with extracts from the original texts that the summary is based on. Going to the sources to verify the AI descriptions is simple, as we are given references for every statement. Here is the summary of a topic about how fossil fuel cars are being phased out, with bans to be introduced in 2030 in several places:
Exploring a topic landscape like this, which we strongly recommend you to try yourself, you will be able to read about planned March landings, projected growth of the markets for AI chips and green hydrogen, challenges related to food insecurity, speculations about what country will get to arrange the 2030 football world cup, and a lot more.
You will probably get new questions while reading about the topics you encounter in the landscape. At that point, you may want to turn to a Dcipher Research Bot. While content landscaping is all about open-ended exploration, the research bots come in when you want help finding answers to specific questions. The two tools complement each other and together form what we call the Insight Booster Toolkit. Let’s train a Dcipher Research Bot based on the same news articles we used for the landscaping and ask it a few questions:
The approach described in this blog post can be used to look at what news or social media have to say about any point in the future—be it next month or year 2100. And you can of course use them to look into anything else as well! For further guidance, take a look at our article on how to make the most of our Insight Booster Toolkit.