
Joe Trippi’s 2004 work to mine the internet and empower supporters of the Howard Dean campaign was revolutionary. As he described it, it was an “open-source revolution” that went beyond the dissemination of campaign messages. Using Meetup.com, blogs, and other media, he turned hundreds of thousands of volunteers into decentralized, self-organizing activists who powered fundraising and local organizing — like a “virtual mid-size city.” It was something to see in real time.
Since then, there have been two distinct iterations in the use of information technology in campaigns. The first was the Republican Party’s use of Cambridge Analytica to microtarget individual voters during the 2016 Trump campaign. While the success of this operation continues to be debated — and how it worked was not transparent — it was a compelling idea for moving beyond bulk messaging that delivers identical messages regardless of individual differences. What made it a game changer was that voter persuasion could be individualized at scale. On the darker side, Cambridge Analytica announced it was shutting down and filing for insolvency in May 2018. The closure was a direct result of intense media scrutiny, investigations, and the loss of clients following the March 2018 revelations that it misused data from up to 87 million Facebook users.
That progressives need to catch up with Republicans in the use of technology seems evident. This challenge is complicated by the advent of readily available, yet still unproven, artificial intelligence technologies like Claude, ChatGPT, and Google Gemini.
Today, it isn’t entirely clear how artificial intelligence will be used in campaigns. We do know a few things. U.S. Senator Bernie Sanders recently sat for an interview with Anthropic’s Claude. (Click here for a clip from that conversation, which exposes some of the motivations for collecting data from internet users.) We also know we need to balance ethical safeguards on AI with innovation in tools that could benefit progressive causes. Finally, misinformation and AI-generated propaganda could undermine democratic processes. What do we do?
What we can’t do is stick our collective progressive heads in the sand. I can’t count how many people I’ve heard say something like, “AI uses too much energy, so I won’t use it.” Two things about this. First, privacy issues are more important than energy use. Second, energy use compared to what?
In her new book, Clearing the Air: A Hopeful Guide to Solving Climate Change in 50 Questions and Answers, author Hannah Ritchie writes, “Data centers currently use only a few percent of the world’s electricity. The big question, though, is whether this will explode with the rise of AI. Probably not.” She discusses a Pareto-style analysis that points to the true energy hogs. Not surprisingly, these are industry, buildings, electric vehicles, air conditioning, and heating, with data centers eighth on the list at around 1-3 percent of consumption. At a minimum, progressives need to stop hyping unknown energy scenarios and instead resolve issues around privacy (Senator Sanders has a bill) while pressuring Google, OpenAI, and Anthropic to meet their corporate climate goals.
Dealing in facts, not hyperbole, is always good advice.
AI is imperfect and no substitute for grassroots knowledge about campaigns and the real voters who will participate in elections. While the database of personal profiles AI draws upon is vast, the granular knowledge that a political activist in a specific race possesses is more relevant to an individual’s potential behavior than AI ever will be.
Like other technologies, AI is a tool that belongs in campaign toolboxes. It is an extension of what Joe Trippi did so long ago — and it is worth learning about instead of shunning.













You must be logged in to post a comment.