📬 Letter from the Editor: On Local Journalism, Power, Tech, and What Comes Next

Dear Reader,

Over the past year, I’ve been thinking a lot about why this newsletter exists, not just what it covers. I began writing the Skamania Dispatch and the Klickitattler in April 2025 after finding that items important to me had passed through public meetings on largely unread, unmentioned consent agendas. The kind of coverage I needed as a resident didn't exist yet, and I knew that I had the tools and the interest to correct some of the imbalances that created this situation.

Skamania County has been described accurately as a local news desert, and Klickitat experiences similar up-to-date news coverage struggles across its wide geography. Like many rural places, we’ve watched local papers shrink or disappear as private equity, hedge funds, and collapsing ad markets hollow out the newsrooms that once attended meetings, read the packets, and asked hard questions on behalf of the public. As Poynter reports:

Medill’s 2025 State of Local News report tracked 136 newspaper closures over the past year, up from 130 last year. In total, the country has lost nearly 3,500 newspapers and more than 270,000 newspaper jobs over the past two decades, leaving 50 million people in “news deserts,” areas where people have limited or no access to reliable local news sources.

From across the country to here in the Gorge, it’s a structural failure, one that leaves communities like ours with fewer eyes on power and public process.

What often gets lost in discussions of “news deserts” is just how time-intensive local accountability reporting really is. Public meetings stretch late into the night, or happen during hours difficult for members of the public with day jobs to attend. Even officials struggle to fully read the agenda packets, which can be hundreds of pages. Decisions unfold slowly across months or years, scattered across boards, districts, and jurisdictions across the Columbia River Gorge. For residents (who often live and work across multiple counties, multiple states) trying to follow along after work, that sheer volume becomes a barrier. Not because people don’t care, but because the system assumes unlimited time and labor.

This is where I believe we’re at a technological turning point.

The advent of recorded meetings in the COVID-era, combined with now accessible technical tools for transcription, document analysis, and cross-referencing offers something genuinely new: a way to rebalance that equation. These tools, including Large Language Models and their AI-backed software implementations, in my opinion do not replace human-generated journalism. They don’t add judgment or context on their own. But they do compress time, and give incredible new firepower to journalists and newsrooms who can use them well and responsibly. They make it possible for say, a single person like me to digest hours of meetings, search dense packets, and surface key decisions without requiring a full newsroom staff or a 40-hour workweek dedicated to just one beat. Let's zoom in on what that looked like for this past 8 months of coverage:

📊 By the Numbers: What This Coverage Took in 2025

📰 Posts published:
• 72 public-meeting summaries (county, cities, districts)

🎧 Meetings "listened" to as transcripts:
• ~350–430 hours of recorded meetings
(about 4–6 hours per post)

📄 Documents reviewed:
• ~7,000–10,000 pages of agendas, packets, staff reports, budgets, and appendices
(often hundreds of pages per meeting)

✍️ Writing, fact-checking, & formatting:
• ~100 additional hours turning notes into readable, shareable, verifiable summaries

⏱ Equivalent labor without LLM-transcription and document-sorting support:
• ~1000 to ~1,370 hours of listening, reading, reporting and review

That’s roughly equal to:
➡️ ⅔ of a full-time journalist’s annual workload or $40,000-60,000 of labor, done by me as a volunteer, providing a free newsletter

Some of the core tools I used, like Downie and MacWhisper, ensured that I did not need to sit through hours of meetings in realtime, but could instead read them and focus my time on portions of meetings with the most content, rather than cross-talk or technical hiccups. While peers express thoughtful reservations about the risks Generative AI poses to journalism, and related threats of misinformation, I believe that when used carefully and shared openly, these AI-backed tools can allow communities like ours to reclaim something that was slipping away: the ability to see what our institutions are doing, while there’s still time to respond. As the Harvard Gazette discusses:

“The question today isn’t whether we are using AI in journalism, because we do it already,” but whether “we can do journalism without outsourcing our skepticism, our ethics, and our sense of accountability, both as journalists ourselves and the accountability we are asking people and organizations that hold power to provide,” said Sideris who is studying how generative AI can better assist investigative reporting as a 2026 Nieman Fellow.

Equally important is how these tools are used, and that they do not work alone. This newsletter exists within, and is made possible by, a rich and deeply appreciated tapestry of Gorge-wide reporting. In-depth coverage from Columbia Gorge Documenters (powered by Uplift Local), Uplift Local, Columbia Gorge News, OPB, KGW, and other regional outlets provide the investigation, narrative, and lived context that meeting summaries alone can’t capture. They provide important sources for cross-referencing LLM-transcribed material to ensure accuracy, and for readers who desire to dive deeper to continue their engagement.

In between I have been proud to share what I've learned with other local journalists, how to use these tools for transcription and sorting documents. I've been excited to teach my community about public records, and how to request them with the help of platforms like MuckRock. We've held free public records workshops, collaborated on longer term investigations, and shared how LLM-backed tools can both help create effective requests and quickly sort through the piles of results. This kind of peer-to-peer learning matters. It’s how capacity grows in places where money is scarce, but commitment is not.

I don’t see this newsletter and any analysis generated with AI tools as a replacement for traditional local journalism. Quite the opposite. It exists in conversation with and deep gratitude for the remaining reporters, editors, and documenters who continue to do in-depth, original reporting across the Gorge under incredibly difficult conditions, with limited budgets. My hope is that this kind of work can help hold the line and create shared civic memory, fill gaps where they exist, and make it easier for others, including future journalists, to step back into these beats.

It's my dream that the next year for the Skamania Dispatch and the Klickitattler will include using these tools to double down on their core mission of community accessibility, like translating this newsletter into a Spanish language edition, mail-based outreach to new neighbors, and sharing content across social platforms like Instagram and Facebook.

If the next chapter of local journalism in rural places like Skamania and Klickitat is going to look different, it will have to be built differently too: with shared tools, shared skills, shared records, and a shared belief that transparency shouldn’t be a luxury reserved for places with large newsrooms and deep pockets.

Paying attention together is not just an act of curiosity. It’s an act of civic self-defense. Thank you for making this newsletter a part of your day, and for being invested in our future together as neighbors.

— Kate