← Back to Library

Use AI this election

This piece flips the script on political apathy by arguing that artificial intelligence isn't a threat to democratic engagement, but the very tool needed to rescue it. Scott Alexander suggests that the barrier to voting responsibly isn't a lack of interest, but the prohibitive time cost of researching obscure local races—a problem AI can solve in seconds. By treating the ballot not as a mystery to be feared but as a dataset to be queried, the article offers a pragmatic path forward for the overwhelmed voter.

The Efficiency of the Centrist Algorithm

The core of Alexander's argument rests on a simple, provocative premise: if you are willing to spend an hour researching a local election, AI can do it for you in minutes, making the effort actually feasible. He demonstrates this by feeding a large language model a detailed persona: "I'm a centrist liberal abundance YIMBY whose favorite political writers are Kelsey Piper, Matt Yglesias, and Ezra Klein." The result is not a generic summary, but a tailored analysis that maps candidates against specific policy preferences.

Use AI this election

Alexander writes, "AI makes it so much faster that you might want to start [researching]." This is the piece's most compelling claim. It shifts the conversation from "should we use AI for politics?" to "how can we afford not to?" The author illustrates this with the California Superintendent of Public Instruction race, where the AI correctly identifies that the office is less about direct control and more about a "$150B-budget bully pulpit." This nuance is often lost in standard voter guides but is crucial for setting realistic expectations.

The analysis then drills down into the "Kelsey Piper view," a specific ideological framework emphasizing structured literacy and the "Mississippi Miracle." Alexander notes that the AI successfully identifies that Piper's project sits at the intersection of positions that "don't all naturally cluster," such as supporting phonics-first instruction while remaining skeptical of teachers' unions when they block reform. This level of granular filtering is exactly what busy voters lack the time to perform manually.

"The structured literacy litmus test points hard to Muratsuchi. He is the only candidate in this field who has actually moved policy on the thing Piper considers most urgent."

By cross-referencing candidate records with specific policy demands, the AI cuts through the noise of campaign slogans. Alexander points out that while some candidates offer vague promises of "more funding," the AI highlights that Al Muratsuchi co-authored the law funding phonics-based instruction. This specific, evidence-based matching is where the technology shines, transforming a ballot full of names into a clear choice based on policy alignment.

Navigating Local Governance and Institutional Trust

The commentary doesn't stop at high-level policy; it tackles the messy reality of local measures, specifically the Peralta Community College District tax reauthorization. Here, Alexander uses AI to balance the "straightforward yes case" against the district's troubled history. The tool explains that the measure is a reauthorization of an existing tax, not a new one, and that funds are restricted to instruction with citizen oversight. Yet, it doesn't shy away from the "accreditation crisis" that saw the district placed on probation in 2020 due to mismanagement and a "broken board culture."

Alexander writes, "I want to give you the honest picture rather than just the campaign pitch." This commitment to transparency is vital. The AI acknowledges that while accreditation was restored and a stabilizing chancellor was hired in 2024, the district is still navigating a budget crisis. This dual perspective—acknowledging the low cost and high utility of community colleges while admitting the institutional dysfunction—provides a more complete picture than a partisan endorsement ever could.

Critics might note that relying on AI to interpret complex local governance risks oversimplifying the nuances of board culture or the specific legalities of parcel taxes. An algorithm might miss the subtle political dynamics of a "secretive meeting" or the human element of a "culture of incivility" that a local journalist would catch. However, Alexander mitigates this by using the AI as a starting point for research, not a final verdict, noting that he ultimately voted for a different candidate than the AI recommended because the tool helped him "prioritize what to research further."

"This is better than any voter guide I've ever seen."

This statement underscores the potential of AI to democratize information access. By synthesizing data from disparate sources—ballot measures, candidate bios, union endorsements, and historical context—the tool creates a voter guide that is both comprehensive and personalized. It allows voters to apply their own values, whether that's a focus on "science of reading" or "school choice," to the specific candidates on their ballot.

Bottom Line

Scott Alexander's piece succeeds by reframing AI from a source of misinformation to a powerful engine for civic diligence. The strongest part of the argument is the demonstration that AI can effectively apply complex ideological frameworks to local races, turning a daunting task into a manageable one. The biggest vulnerability remains the reliance on the quality of the underlying data and the risk of algorithmic hallucination, but the author wisely positions the tool as an aid to human judgment rather than a replacement. For the busy voter, this is a compelling case to stop scrolling and start prompting.

Deep Dives

Explore these related deep dives:

  • Mississippi Miracle

    The article cites this specific turnaround in reading scores as the empirical foundation for Kelsey Piper's argument that structured literacy and phonics-first instruction are effective, countering the prevailing skepticism in California.

  • Nonpartisan primary

    Understanding this unique California electoral system is essential to grasping why the article describes the Superintendent race as a standalone contest with a potential runoff, rather than following the standard partisan primary format.

  • Science of reading

    This technical debate over cognitive science and literacy instruction is the core ideological fault line the author uses to distinguish Kelsey Piper's evidence-based approach from the traditional union-backed methods she critiques.

Sources

Use AI this election

by Scott Alexander · Astral Codex Ten · Read full article

I’m not saying AI is superintelligent or can decide better than you can. I’m saying that if you - like me - spend an hour or so doing research before voting on local seats, AI can aid that research very effectively. And if you don’t do that research - because you weren’t willing to waste an hour on it before - AI makes it so much faster that you might want to start.

I gave Claude a prompt something like (edited for coherence):

I’ll be voting in the June 2026 California primary. I’m a centrist liberal abundance YIMBY whose favorite political writers are Kelsey Piper, Matt Yglesias, and Ezra Klein. I’m wary of government overreach, but I’m not a doctrinaire libertarian and want to help people when we can figure ways to do it that work. I’m going to ask you about each race on my ballot, and I’d like for you to list the various candidates’ bios, policies, endorsements, your read on the most important differences between them, and your advice for me as I try to make my choice.

…and got back answers like the following:

California Superintendent of Public Instruction.

A quick reality check on the office, then the field. The state superintendent has surprisingly little direct authority — districts control most budget and curriculum — and the state is actively considering shifting more responsibilities to the governor-appointed State Board of Education, which would shrink it further. The job is mostly a $150B-budget bully pulpit. So this is more “who do you want amplifying education priorities for the next four years” than “who will rewrite California education.”

The race is on the standalone education ballot (it doesn’t follow the nonpartisan blanket primary system) — June 2 primary, November 3 runoff if no one clears 50%. Realistically no one does, so this primary is about who advances.

Quick framing: what “the Kelsey Piper view” implies.

Piper’s project, especially in her recent Argument writing, sits at the intersection of several positions that don’t all naturally cluster:

Structured literacy / “science of reading” / Mississippi Miracle is real and works. Phonics-first, mastery-based, with district-level accountability and reading coaches in low-performing classrooms. She has called the failure to adopt this “a crime against humanity” in California.

Standardized testing and accountability are good, actually. Not as bludgeons but as honest feedback mechanisms. Detracking failed. Grade inflation is bad.

Open to charters, microschools, ...