AI ON THE FRONTIER

Practical AI insights for Cochise County nonprofits and educators

Issue 4  •  July 2026

Hello! July arrives in Cochise County with triple-digit heat and, this year, a pair of posts that ask a question that does not get asked often enough about artificial intelligence: who decides how it gets used, and who bears the cost when those decisions go wrong? This month's deep dive reaches back to the early 1800s for an answer that turns out to be surprisingly current. The second post looks at how one of the largest technology companies in the world may be making that question urgent for anyone who uses a PC.

Deep Dive

The Wrong Target: What the Luddites Can Teach Us About AI

We use the word "Luddite" to dismiss anyone who pushes back on new technology, but the actual Luddite movement had almost nothing to do with opposing machines. In the early 1800s, English textile workers watched as mill owners introduced machinery to undercut wages and replace skilled labor with no transition support and no consideration for what happened to the workers left behind. The machines became the target because they were visible. The mill owners who decided to deploy them that way were not.

That same dynamic keeps appearing in the AI conversation today. When a company reduces headcount because AI now handles functions that employees used to perform, the backlash lands on the AI. But the technology did not make that decision. Most people who are uncomfortable about AI are not actually opposed to it in principle. Nobody is protesting hospitals for using AI to read imaging scans. What they are reacting to is a specific, repeatable pattern: AI deployed as a cost-cutting mechanism with no transition support and no acknowledgment that affected workers have legitimate grievances. That is a reasonable response to a human decision about how a powerful tool gets used.

For smaller organizations introducing AI tools internally, the lesson is about how. Transparency about what you are doing and why, genuine involvement of the people who will be affected, and honest conversation about what will change are not just courteous gestures. The mill owners were not wrong that the machines were more efficient. They were wrong in their indifference to the people who paid the price for that efficiency. AI is not the enemy of workers. The indifference of decision-makers to the workers AI displaces is. And unlike the machines themselves, that indifference is something we can actually address.

Read the Full Article →

What's New at Cochise AI

▶ Microsoft Is Making Copilot Inescapable. That's Not the Same as Making It Good. Microsoft has pressed Copilot into Windows, Office, physical keyboard hardware, and now a wearable AI badge announced at its Build 2026 conference. The concern is not Microsoft per se but a pattern the tech industry has seen before: when a dominant platform vendor picks a winner by controlling distribution, quality becomes largely irrelevant to market share. The same thing happened with Internet Explorer in the late 1990s, and browser innovation stagnated for nearly a decade as a result. Millions of people may form their understanding of what AI can do based on the least impressive version of it, simply because it came pre-installed. Before you default to whatever AI tool arrived with your existing software, try the task that actually matters to your organization in two or three different tools and compare results honestly.

Worth Knowing

▶ Apple Rebuilt Siri from the Ground Up. At its WWDC 2026 developer conference, Apple replaced the old Siri engine with Google's Gemini AI and gave it a dedicated app with persistent conversation history that syncs across iPhone, iPad, Mac, and Apple Watch. If Siri has been a frustrating non-starter for your staff compared to ChatGPT or Claude, that may be changing. Apple says conversation history stays private through iCloud, which matters for anyone using it with work-related content.
▶ Real-Time Voice Translation Is Now in Your Pocket. Google's Gemini 3.5 Live Translate handles real-time speech-to-speech translation across more than 70 languages, available now in Google Translate on Android and iOS. For nonprofits and schools serving Spanish-speaking families or other multilingual communities in Cochise County, this is worth testing. It eliminates the stop-and-wait pattern of typed translation and works well enough for genuine back-and-forth conversations.
▶ An AI Filing Mistake Just Cancelled a Court Trial. A Mississippi federal court cancelled a trial after discovering that lawyers on both sides had submitted filings containing AI-generated errors. The lesson is not that AI is unusable for writing. It is that anything going out officially under your organization's name — grant applications, board reports, legal correspondence — needs a person to review it who can catch factual errors and verify that every claim is accurate. AI drafts well. It does not proofread itself.
▶ Anthropic Released Its Most Capable Model Yet. Anthropic launched Claude Fable 5, which independent reviewers describe as a meaningful leap over previous versions in writing quality, reasoning, and complex task handling. If your organization uses Claude through Claude.ai, you likely already have access to the new model at no additional cost. It is worth re-trying tasks where Claude's responses have felt incomplete or shallow.

Cochise AI, LLC

Sierra Vista, AZ  •  cochiseai.com

Want to receive this newsletter by email each month?
Sign up here