In just the past few months, artificial intelligence has gone from a toy to a practical political campaign tool. Campaigns, political organizations and consultants are using tools like ChatGPT to create email fundraising messages, cut walk lists for volunteers, analyze demographic data, produce memes, crank out press releases, conduct opposition research and much more.
While the technology is powerful, the current generation of AIs aren’t exactly a panacea. Chatbots mimic human communications, for example, but they don’t understand it. They can make fundamental mistakes or even invent facts out of whole cloth. AI may help us get more work done in a given amount of time, but we have to build in safeguards to make sure that information it gives us isn’t from an alternate universe.
I explored these questions in depth in this month’s Campaigns & Elections column, which is entirely devoted to AI and machine learning in politics. I look at “training” AIs to get better results, along with a look into just what might go wrong if you take an AI’s analysis at face value. Check out the full article, and let me know how you’re using AI in your own political work. While the machines still let us run our own affairs….