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, and school choice. She founded and runs Oakland LEARN, a microschool. She sends her own kids to an “unregulated tiny private school.” She thinks the left’s reflexive hostility to choice misunderstands what parents are actually fleeing (kids not learning to read).

Skeptical of teachers’ unions when they block reform. Not anti-teacher, but pro-evidence over pro-union-priority when they conflict.

Pro-public-school in principle, but pro-evidence over pro-system. Treats education as a place where the left has often chosen ideological commitments over the well-being of kids who do not vote.

The candidate who matches this view best probably (a) has explicitly backed structured literacy/phonics, (b) has shown willingness to push the system rather than just defer to CTA, and (c) is at least not hostile to charters or alternative models. The candidate who matches it worst would be a CTA-backed insider running on “more funding, less testing, no curriculum mandates.”

The candidates

Al Muratsuchi (Assemblymember, LA South Bay, former Education Committee chair). Co-authored the law that funded phonics-based reading instruction grounded in the science of reading. Co-authored Prop. 2 ($10B school facilities bond, passed). Adjunct college teacher, former deputy AG. Endorsed by California Federation of Teachers (CFT, the smaller of the two teachers’ unions), Association of California School Administrators, and California School Employees Association.

Anthony Rendon (former Assembly Speaker 2016–2023). Ran an early-childhood program in LA for 20 years before politics. As Speaker oversaw major K-12 funding increases, universal transitional kindergarten, and charter school accountability legislation (which generally tightened charter regulation rather than loosened it). Endorsed by AFSCME, SEIU California, California Latino Caucus.

Joshua Newman (former state senator, former Senate Education Committee chair). Co-authored Prop. 2 with Muratsuchi. Army veteran, founded a vets-to-civilian-workforce nonprofit, now a UC Irvine fellow. Endorsed by State Building and Construction Trades.

Richard Barrera (San Diego Unified school board president). The California Teachers Association candidate — that’s basically his profile. Outsider to Sacramento; emphasizes funding, early childhood, easier teacher credentialing pathways.

Nichelle Henderson (LA Community College trustee, longtime teacher). Union-backed. Supports expanded school-based health, standardized testing reform, divesting pensions from fossil fuels and “the U.S. war machine.”

Frank Lara (SF elementary teacher, former UESF VP). The left-of-the-Democrats candidate. Peace and Freedom Party and Green Party endorsements (officially nonpartisan, but signals where he sits).

Gus Mattammal (Silicon Valley tutoring company founder, author). Pro-charter, pro-homeschool, pro-alternatives-to-traditional-public-schools. The closest thing to a school-choice candidate in the field.

Sonja Shaw (Chino Valley School Board president). Anti-trans-protections crusader who got Thurmond escorted out of a 2023 board meeting. CA Republican Party, Moms for Liberty, CRPA endorsements. The right-culture-war candidate.

The most important differences, and my read for the Piper view

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. His official candidate statement explicitly cites “phonics-based reading instruction based on the science of reading” as something he coauthored and delivered funding for. That is, almost verbatim, the policy Piper has spent the last year writing about. Nobody else in the field has a comparable record on this issue.

Rendon is the second-strongest plausible fit, but with caveats. Twenty years running an early-childhood program is genuinely relevant experience, and TK (transitional kindergarten) expansion is one of the more evidence-supported things California has done in K–12. But his charter-accountability work as Speaker generally meant more regulation on charters, which cuts against the Piper-flavored openness to alternative models. He’s a competent establishment Democrat; he’d be a fine superintendent, just not specifically the structured-literacy crusader.

Barrera is the candidate Piper would be most skeptical of. Not because he’s a bad person — he’s a long-serving school board president, presumably competent — but because the CTA endorsement is his political identity in this race, and Piper’s whole thesis is that CTA-aligned politics has been one of the structural reasons California can’t adopt the reforms that worked elsewhere. Barrera’s stated priorities (more funding, easier teacher pathways) are CTA priorities, not reform priorities. If you’re voting the Piper view, this is the one to skip.

Henderson and Lara are further from the Piper view. Henderson’s “standardized testing reform” plus the pension-divestment-from-the-war-machine framing reads as movement-left rather than evidence-driven; Piper has been pointedly critical of detracking and of treating testing as the enemy. Lara is more or less explicitly the union-left candidate.

Mattammal is the awkward case. He is the candidate closest to Piper on school choice and microschools specifically — that’s basically her own model. But he’s running with essentially no political coalition or governing experience, and the Piper view, while choice-friendly, is not “burn down public schools” — it’s “fix the actual reading instruction in the schools that most kids will continue to attend.” A statewide superintendent who can’t move CTA-dominated Sacramento isn’t useful for that goal. Mattammal also doesn’t appear to have an articulated policy program on structured literacy or accountability, just on alternatives. Probably not the right fit even on a choice-sympathetic reading.

Shaw is obviously off the table on the Piper view. Piper is a Bay Area liberal who supports trans kids and has no interest in culture-war education politics. The Moms for Liberty endorsement is disqualifying on its own.

Newman is the dark horse. Co-authored Prop. 2 with Muratsuchi, chaired Senate Education, military background, now an academic fellow rather than a Sacramento insider. He hasn’t been as visibly identified with structured literacy as Muratsuchi has, but he’s broadly in the same evidence-and-accountability lane. The Building Trades endorsement is unusual — it suggests he’s running a coalition strategy aimed at facilities/jobs rather than CTA — and that might actually be a positive signal for someone trying to operate independently of the dominant teachers’ union politics.

Bottom line

For the Piper view, Muratsuchi is the clearest fit in the field. He’s the only candidate who has actually legislated structured literacy funding, he has the directly relevant policy chops (former Education Committee chair, Prop. 2), and his endorsement coalition (CFT, administrators, classified employees) is just slightly off-axis from the dominant CTA-controlled lane in a way that suggests some independence. He’s also reasonably likely to make the November runoff, which matters: a vote for the second-best candidate who finishes fourth doesn’t accomplish anything.

If you wanted to express a sharper “I want a real outsider reformer” view rather than working within the field as it exists, Mattammal is the only option, but at significant cost in viability and breadth of program.

For “I just want a competent Democrat who’ll be fine,” Rendon and Newman are both reasonable. But for “I want the candidate most likely to actually push the structured-literacy / accountability agenda Piper writes about,” it’s Muratsuchi.

This is better than any voter guide I’ve ever seen. I didn’t end up taking Claude’s advice - I voted Newman - but Muratsuchi was a close second, and Claude’s advice was crucial in helping me prioritize what to research further.

And here’s its opinion on one of the small local spending measures I usually have a hard time forming an opinion on:

Measure A: Peralta Colleges Affordable Education Reauthorization

What it actually does

This is a parcel tax reauthorization, not a new tax. The Peralta Community College District (which runs Laney College, Merritt College, College of Alameda, and Berkeley City College) has had a $48-per-parcel annual tax since voters first approved it in 2012 and renewed it in 2018. The current authorization expires.

Measure A would:

• Extend the existing $48/parcel tax for 9 more years (July 1, 2027 through June 30, 2036)

• No increase in the rate — same $48 you’re already paying if you own property in the Peralta district

• Generate roughly $8 million per year ($72M over the nine years)

• Funds restricted to instruction, student services, counseling, faculty recruitment/retention, job-training programs, and university-transfer prep

• Requires a 2/3 supermajority to pass (this is the standard threshold for California parcel taxes and is a high bar)

Senior citizens and Supplemental Security Income recipients can apply for exemptions. Funds are subject to a Citizens’ Parcel Tax Oversight Committee.

Geographic scope: this only appears on ballots in the Peralta district, which covers northern Alameda County — Oakland, Berkeley, Alameda, Albany, Emeryville, and Piedmont. So if you’re in Oakland, you’ll see it.

The straightforward “yes” case

• Community colleges are genuinely one of the most cost-effective rungs in American higher ed. Peralta tuition is ~$1,100/year vs. ~$6,500 at Cal State and ~$14,000 at UC.

• Reauthorizing an existing tax is a much lower ask than a new one — you’re already paying it, and nothing on your tax bill changes.

• The money is restricted to instruction and student services (not administration, not facilities), with citizen oversight. That’s about as well-designed as parcel tax measures get in California.

• League of Women Voters supports it. Labor unions back it. No formal opposition argument was filed with the Registrar of Voters — literally no one wrote a ballot statement against it.

• $48/year on a parcel is genuinely small relative to most California property tax bills.

The complications worth knowing

I want to give you the honest picture rather than just the campaign pitch, because Peralta has a difficult institutional history that someone with your “wary of government overreach but pro-helping-people” framework will want to understand.

The accreditation crisis was real. From 2018 to 2023, Peralta was in serious trouble with its accreditor (ACCJC). A 2019 state Fiscal Crisis and Management Assistance Team review found Peralta was “at high risk of insolvency” after years of mismanagement, with administrators “regularly breaking their own rules” and the district “ripe for fraud” due to poor internal controls. The four colleges were placed on probation by the accreditor in January 2020 — a severe sanction one step short of “show cause” (the level that nearly closed City College of San Francisco).

The governance situation was also a mess. A 2021 Alameda County civil grand jury report excoriated the Peralta Board of Trustees for “broken board culture,” secretive meetings, infighting, interfering in hiring, racial insensitivity, and fostering an “overall culture of incivility.” Multiple chancellors departed under pressure — Peralta went through four chancellors between 2019 and 2023.

Things have improved. Accreditation was restored from probation to “warning” status in January 2022, then fully reaffirmed in January 2023. Tammeil Gilkerson became permanent chancellor in January 2024 — by all accounts a stabilizing hire from outside the dysfunction.

But there are ongoing concerns. As of early 2025, Peralta is still navigating a budget crisis. Chancellor Gilkerson has acknowledged that Peralta’s full-time-equivalent students (FTES) lag significantly behind other four-college community college districts in the state — which means enrollment-driven state funding is structurally low for the district’s size. There’s also a Peralta Citizen article headlined “PCCD is in crisis. Everything is on the table” from February 2025, indicating that even with the recent improvements, the district isn’t out of the woods.

One transparency concern worth noting from local reporting. The largest single donor to the Yes-on-A campaign is Future Flooring Group (C&S Flooring Systems), which contributed $75,000 — about half the campaign’s total raised. Future Flooring is currently contracted with Peralta for ~$678K to replace flooring at College of Alameda, Laney, and the district service center this summer, and has received over $2M from the district overall. This isn’t illegal, but a major vendor that depends on district contracts being the largest single funder of the campaign to extend the district’s parcel tax is the kind of thing that should at least be visible. The rest of the campaign funding comes mostly from Bay Area labor unions, which is more typical and less notable.

My honest read for your framework

For a centrist liberal abundance YIMBY who wants to help people but is wary of government overreach, here’s how I’d think about this:

Arguments for yes:

• Community colleges are exactly the kind of institution that fits an “abundance liberal helping people” framework: they provide cheap, accessible higher ed and workforce training without the racket-like price escalation of four-year schools. They’re a genuine social ladder.

• This is a reauthorization, not new spending. Voting no doesn’t return the money to you in a meaningful way at the household level; it just defunds the colleges.

• $48 a year is small money. The cost-benefit is hard to argue against in isolation.

• Peralta has measurably improved governance since 2023. Punishing the institution now for sins it has been actively repairing would be unfair to current and future students.

• The supermajority threshold (2/3) is itself a check — if a tax can clear that bar, it has unusually broad consent.

Arguments for no:

• Peralta’s institutional dysfunction was severe and is not fully resolved. The “everything is on the table” framing from district leaders in early 2025 suggests structural problems persist.

• Some abundance-school thinkers would argue that propping up underperforming public institutions with parcel taxes reduces pressure to actually fix what’s broken — if the colleges had to demonstrate enrollment growth and operational excellence to retain funding, they might.

• Low FTES is a worrying signal. Are these colleges serving the demand that exists, or are they an institution that has become inefficient relative to peers?

The judgment call. I think a reasonable abundance liberal would vote yes, with eyes open. The reasons: (1) reauthorizations are different from new authorizations — you’re not signing up for new spending, you’re declining to actively cut it; (2) community colleges as a category are deeply aligned with the “help people, expand the ladder” framework; (3) the most damning Peralta criticisms are about 2018–2022, and the institution has been visibly improving since; (4) no organized opposition emerged, which usually means even motivated critics didn’t think the case for opposing was strong; and (5) $48/year is genuinely small money for a meaningful local good.

But voting no isn’t crazy if your prior is “underperforming institutions should face budget discipline.” It’s an unusually defensible “no” for a parcel tax, in that the institution has a real track record of problems. If you wanted to send a message that you support community colleges in principle but want Peralta to keep cleaning house, abstaining is also coherent (the 2/3 threshold means abstaining has a slightly different effect than a no vote — it raises the bar relative to the participating electorate).

My weak lean: yes. Strong enough to recommend, not strong enough to argue with someone who voted no on accountability grounds.

I informally (no strict controls) tested this on 10 difficult races where I wasn’t sure going in who I would vote for. I give Claude an A if it recommended the same candidate I eventually voted for, a B if it recommended my second choice out of a large field, a C if I can see where it’s coming from but disagree, and an F if it picked someone I hate. Its final score was:

• 5 x As

• 3 x Bs

• 2 x Cs

When it failed, it was usually because I have a policy of making my vote count on important races, vs. making fun/utopian/conscience-guided choices on races that don’t matter. I didn’t tell Claude this, and Claude tended to ignore fringe candidates with no chance of winning.

Other times, it was because it took my prompt too seriously. I do like abundance/YIMBY ideas, but since that was all I told Claude in the prompt, it sort of pegged me as a single-issue voter. In the future, I might put that lower down, below the list of famous people I agree with.

Claude also correctly guessed (probably from the word “liberal”) that I lean Democrat, over-indexed on that, and leaned away from Republican candidates. In the future, I might include in the prompt that it should cross the aisle when there’s a sufficiently convincing reason.

I ended by asking Claude to ignore my personal beliefs and tell me who it would recommend, point-blank, if it didn’t know anything about me. It proposed four switches - three of which I liked better than its original choice! I’m skeptical that it was truly expressing its own opinions here (if it even has any) - I suspect it treated the exercise as a chance to make more “emotional” judgments - but it was an interesting exercise. If you’re a Republican using AI to help you vote, please end with the same question and tell me what happens.

This was a good enough experience that if I didn’t have an hour or two to do this properly, I would trust Claude’s endorsement over competing cheap-and-fast ways to determine my vote (party-line, newspaper endorsements, NGO endorsements, who has the funniest name). And given that I did spend two hours doing this properly, I think I’m about 50% happier with my choices than I would have been without AI.

Moving back into speculative futurism (in case you forgot which blog you were on), a lot of my hope for how the post-AGI future might turn out well - despite all the risks of dictatorship, inequality, surveillance, etc - hinges on people taking advantage of AI advisors to make better political decisions. The sooner we start, the better.

You can see my final list of decisions (may not be very interesting to non-Californians) here. Remember that I accept voter guides from ACX meetup groups in big cities with local elections happening soon, and will signal-boost them on the Open Thread if you give them to me.