Using Deep Research Versus Party Loyalty

Share Copied!

In this episode

The connection

You asked about who to vote for in the upcoming midterms, a question that usually means facing a mountain of glossy mailers and loud TV ads. Writing in May, Scott Alexander suggested a way to bypass the shouting by using AI tools to find local candidates who share specific, personal values. It’s a picture of the voter as a precise researcher. But just last month, Charlie Hunt pointed to a different reality, explaining why people often ignore that kind of research and stick with scandal-prone candidates just to keep the other side from winning. Something interesting emerges when the desire to be an independent thinker hits the heavy pull of staying loyal to a team.

Scott Alexander envisions a world where a voter can feed their favorite writers into a chatbot and get back a candidate tailored like a custom suit. It’s a hopeful take on what happens when the grunt work of local research disappears. The assumption is that if people could easily see past generic party labels to find a specific "science of reading" advocate or a "YIMBY" ally, they would choose the person who actually fits their life. But move from the local school board to the high-stakes Senate races in Texas or Maine, and that logic starts to fray. Charlie Hunt, Associate Professor of Political Science, Boise State University, points to candidates like Ken Paxton and Graham Platner—men carrying enough baggage to sink a typical career—who still sail through their primaries. Where Alexander sees an opportunity for precision, Hunt sees a blunt instrument. The interesting thing is how these two views sit together. Alexander’s AI tool works best in the quiet corners of government where the "other side" feels less like an existential threat. Once a race hits the national stage, Hunt shows that the desire for a perfect policy fit often collapses under the weight of just needing to win.

A voter sits at a kitchen table with a ballot, using a custom AI tool to find a candidate who matches their exact views on local zoning. It feels like a win for thinking for oneself. But then comes the pull of the party line—the fear that a "pure" vote might just help the other side win. It’s a choice between a tailored fit and a heavy shield. How much of what a person actually believes are they willing to trade away just to keep the wrong side out of power?