American Voters, Large Language Models & Shaping Realities
Our realities are increasingly shaped by what we are “learning” from AI-based sources like ChatGPT, Claude, and Gemini. LLMs (AI Large Language Models) have become part of everyday life and increasingly shape what information people encounter and in what order, subtly deciding what's important and even what is true, sometimes without users realizing it.
The key question is: How are these models trained? Who and what is training the trainers?
That was the topic of discussion for our first 2026 Webinar Series: A New Look at the American Voter this month.
We were joined by Tim Chambers and Mary Coller Albert of Dewey Square Group; Jessica González, the co-CEO of the media and tech public policy organization, Free Press; and Kelsey Suter, a Partner at Upswing Research.
Understanding American voters requires understanding how their opinions are shaped. LLMs play an increasing role in determining how people understand and process information. We ignore this at our peril.
Download Dewey Square Group’s Slides