My friend Katie Harbath has had a front-row seat to nearly every chapter of the collision between technology and politics. Before founding Anchor Change, she built GOP.com, ran digital for Rudy Giuliani's campaign, led digital at the NRSC, and then spent a decade at Facebook running global elections policy. Her new book, Disrupting Politics: A Front Row Seat to the Collision of Technology and Democracy, tells that story from inside the rooms where the decisions got made. On this episode, I talk with Katie about what the book covers and what we should do with its lessons.
We start with a question I've been chewing on: what happened to the high hopes we had for the internet in politics? Katie's answer begins with something mundane: analytics. The first time she could show an RNC press secretary real-time numbers on who was clicking, the incentives changed. More clicks were better, so campaigns did more of whatever got clicks. Over time, like the proverbial boiled frog, that pulled everyone toward the extremes. Add online fundraising that freed insurgent candidates from the party infrastructure, and you get politicians playing to a national audience instead of their constituents.
We also dig into what it was like being the Republican hired to bridge Facebook and the political world, from bringing the Young Guns to campus to the 2016 trending topics controversy and the morning after the 2016 election, when the thing she'd spent 15 years building turned villainous overnight. I share my own whiplash from the campaign side: tactics praised as brilliant when the Obama campaign used them in 2012 were treated as sinister by the Cambridge Analytica era.
Then we turn to AI. Facebook had no political policies for years; AI companies started out restrictive and are now twisting themselves into pretzels deciding what to open up. Katie's advice to them: stop asking how to keep everyone from getting mad, and decide where you're willing to withstand pain.
Katie closes with the lessons she ends the book on: good intentions are not safeguards, scale is not a strategy, and panic responsibly. We can't copy the social media playbook for AI, but we can write the new one with the lessons we wish we'd known the first time.