Everyone can use AI. I help you avoid the pitfalls.

These systems learn patterns from enormous amounts of text, and they can’t reliably tell you where a particular answer came from. Treat what comes back as a draft, never as a source. You are the human in the loop, and that job doesn’t transfer.

Assume anything typed into a free consumer tool may be used to improve that tool unless the terms say otherwise. If you handle client, patient, student, or donor information, settle that question before anyone picks a product, not after. If you’re on a paid or business tier, check the account’s data or privacy settings and turn training off: most paid plans let you opt out, which means the tool can’t use what you type into it to train future models. It’s usually off by default on business plans and on by default on free ones, so check rather than assume.

It will change which parts of a job take the most time long before it replaces anyone. These tools handle drafting, summarizing, and generating options. The parts requiring judgment, relationship, and accountability are the parts they can’t do, usually the parts that matter most.

Detection software is unreliable and produces false accusations, disproportionately against multilingual writers and neurodivergent students whose natural writing patterns (formulaic phrasing, unusual sentence rhythm) get flagged as “AI” by these tools. What works is assignment design plus clear, assignment‑level policy, and asking students to reflect openly on how they used these tools instead of driving the use underground.

It depends heavily on which tool. Education platforms with district agreements handle student data very differently from consumer chatbots, and most have age minimums families never see. I walk groups through what to ask your school.

That risk is real and it’s the one I take most seriously. The answer is learning when to use these tools, when not to, and how to evaluate what comes back, a teachable skill, not an instinct. Pretending the tools don’t exist guarantees people use them badly.

That’s a legitimate position and you’re welcome in the room. AI is not one size fits all: your field, context, and purpose all matter. I only ask for a real rationale rather than a reflex, and a willingness to hear what others have found. Some of my best conversations have been with skeptics, and I’ve changed my own thinking because of them.

With shared understanding and a written policy, which cost nothing. Most strong tools have free tiers, and a group that agrees on what it wants from AI will spend far less than one that buys first and asks later.

Sessions start at $250. Library, nonprofit, and civic‑group rates are different from corporate ones, and travel is billed separately. Tell me about your group and I’ll give you a straight number before you have to convince anyone else.