AI Finally Met the Ballot Box

Hey Everyone - Hope you had a great weekend. This week's story is my favorite kind: the biggest AI news of the week had nothing to do with a model. It happened in a packed gym in Festus, Missouri, and at ballot boxes in North Carolina, and it might be the first real evidence that ordinary people still have a lever on how this technology gets built. Let me walk you through it.

This week:

Let's dive in.

In Festus, Missouri, the city council approved a $6 billion data center by a vote of 6 to 2. The meeting was held in front of a packed gym of residents who showed up to oppose it. The council voted yes anyway. Then the election came, and voters removed every incumbent who had voted yes and sought reelection. One of them summed it up plainly. The vote cost him his seat.

In Stokes County, North Carolina, two commissioners lost their primaries after residents organized around their data center votes, and the county takes the issue up again on July 13, the same day this newsletter lands in your inbox. Warren, Maine held a special town meeting on a moratorium on July 8. In Missouri, the backlash has spilled into Republican primaries, with candidates now competing over who will be tougher on data centers. This is not a partisan story. Prism's headline called it the one thing a divided country agrees on.

The tracking data says these are not isolated anecdotes. Data Center Watch counted 75 projects worth an estimated $130 billion blocked or delayed by local opposition in the first quarter of 2026 alone. That matches the entire previous year of pushback, compressed into ninety days.

I think this is the most important AI story of the week, and it contains no model release, no benchmark, and no funding round.

For three years, the AI story has moved in one direction. The labs decide what to build. The markets decide what it is worth. Washington holds hearings. And the job of an ordinary person, the job this newsletter exists to help with, has been to adapt to whatever arrives. Nearly every Signal I have written shares that assumption. The wave comes, and you learn to surf it.

This is the first story where the influence runs the other way at scale. The AI buildout is not made of tokens. It is made of land, water, power, and permits, and every one of those runs through a room where ordinary people still hold direct power. Not the metaphorical kind. The kind that removes you from office. A packed gym in a town of twelve thousand people just imposed a consequence that no congressional hearing has managed to produce.

There is a familiar trap underneath it, and longtime readers will recognize it. The councils that approved these projects were not corrupt. They were optimizing for what fit in the spreadsheet: projected tax revenue, promised jobs, a headline investment number. The costs that did not fit the spreadsheet turned out to be the ones voters actually vote on: water draw, rising power bills, noise, and the feeling of not being asked. The deal passed the analysis and failed the town.

There is a financial tell underneath the politics too. A J.P. Morgan analysis finds that 60 percent of the data center capacity promised by 2027 has not broken ground. The buildout on paper is much larger than the buildout in the ground, and local consent is quietly becoming a constraint the financial models never priced.

I want to be careful here. This is not a celebration or a lament. Whether a data center is good for a specific town is a real question with real tradeoffs. The tax base and the jobs are not fake. Neither are the water tables and the utility bills. Reasonable neighbors land on both sides. The point of this Signal is not which side is right. The point is where the decision is actually being made, which is not in Washington and not in San Francisco. It is at the county commission, the zoning board, and the town meeting.

For months this newsletter has argued that local community is where trust survives the AI era. The update this week is that local community is also where leverage survives. In a national election, you are one vote among 150 million. In a room of forty people deciding a land use question, you are a meaningful fraction of the outcome, and every commissioner in that room has now heard about Festus. Showing up was always a social hedge. As of this year, it is real power over how AI actually gets built, whichever side of the question you are on.

This past week I put together a private event in Austin for some of the top creators in the city. Small, invite-only, and specifically for people who are actively building and monetizing their businesses. No spectators, no pitch fest. Just a curated room.

The value was obvious within the first hour. When everyone in the room has skin in the game, you skip the small talk entirely. People compared numbers, traded playbooks, and started conversations that I'm confident will turn into deals and collaborations. I've been to a hundred networking events that produced less than this one night did.

The lesson I keep relearning is that the leverage isn't in having the biggest audience. It's in curating and creating value for a specific set of people. An audience is a broadcast. A network of peers is a compounding asset. The momentum and opportunity that come out of one well-curated room can move mountains, and I don't think I'm exaggerating.

This is also the third lane of my distribution strategy. Ads and cold email scale reach, but the network lane is the only one that compounds trust, and trust is the constraint on everything else I'm building.

So I'm going to keep running this play. The next versions will be rooms for community builders and for entrepreneurs, same format: a tight filter, a clear reason to be in the room, and value designed for the specific people invited.

And if you read the Signal above, you'll notice it's the same physics. Small rooms, high leverage. Festus proved it at the ballot box. I watched it happen around a dinner table.

The skill this week is deliberate curation. Keep a living, written list of the people you most want in your orbit, and maintain it the way you would maintain any asset.

The filter is three questions. Do you leave time with this person with more energy than you brought? Would you trust them in a real problem, with money or reputation on the line? Could you build something together if the chance appeared? Anyone who clears all three belongs on the list. Most people clear one.

Then the practice. Write the names down, somewhere between twenty and fifty of them. Review the list quarterly: who is missing, who no longer clears the filter, who you have not spoken to in too long. And use it. When you host something, invite from the list. When an opportunity crosses your desk that is not for you, route it to someone on the list. When you make an introduction, pull both sides from it. A list you never act on is a contact export, not an asset.

Here is why this is an AI-era skill and not just a networking tip. AI has made reach free. Anyone can now generate outreach, followers, and connection requests at infinite scale, which means an unfiltered network is indistinguishable from noise. The scarce thing is no longer access to people. It is judgment about which people, and judgment applied to people cannot be delegated to a model, because the model has never sat across the table from any of them.

One more room belongs on your list, and it is the one this week's Signal is about. Your local meeting, the county commission, the school board, the zoning board, has no curation at all. The filter is simply showing up, which is exactly why the people who show up hold so much leverage. Put one local meeting on your calendar this month and sit in the back. You will learn more about where power lives in your town than any feed will teach you.

Like every skill in this series, this one holds up in every version of the AI future. If the boom continues, your curated room compounds opportunity. If it stalls, your curated room is the safety net. The list works either way.