It predicts, it doesn't know
A model is guessing the next likely words, confidently. Treat every answer as a first draft from an eager new hire, not a verdict.
AI, explained like a friend would
Smarter than the average bear? Nope — and that's kind of the point.
I'm Louise, a non-technical, non-coding citizen developer learning AI out loud. Real experiments, honest gotchas, plain English — so you can skip the intimidation and get to the useful part.
// pipeline integrity by day · power platform tinkerer · figuring it out like everyone else
Welcome to the den ·AI stuff that actually helps ·No jargon, promise ·Written by a human still figuring it out ·Citizen developer notes ·Welcome to the den ·AI stuff that actually helps ·No jargon, promise ·Written by a human still figuring it out ·Citizen developer notes ·Welcome to the den ·AI stuff that actually helps ·No jargon, promise ·Written by a human still figuring it out ·Citizen developer notes ·Welcome to the den ·AI stuff that actually helps ·No jargon, promise ·Written by a human still figuring it out ·Citizen developer notes ·
About Louise
An average bear, learning in the open.
I'm a senior employee at a large pipeline company in Edmonton, Alberta. My work leans toward process efficiency and pipeline integrity — the unglamorous, deeply practical side of getting things done well.
Over the last five years or so I've fallen into the Power Platform world, and more recently I've been digging into AI. I'm self-taught, and I take these tools on as a non-technical person.
The train has left the station on AI. I share what I'm learning to help other average bears upskill, make sense of it, and actually leverage the technology.
Citizen developerPower PlatformProcess efficiencySelf-taughtEdmonton, AB
Read the posts on LinkedInBear necessities
The handful of things worth knowing first
No prerequisites, no acronyms. These are the ideas that make everything else about AI click into place.
Context is the whole job
Goal, audience, constraints, and an example of "good" will do more for your output than any clever phrasing trick.
It has no memory of you
Every chat starts amnesiac. If it matters, say it again — background, tone, the thing you decided last Tuesday.
It wants to please you
Flattery is a feature, not feedback. Ask it to argue against your idea if you want something honest back.
Start where mistakes are cheap
Trip planning, meal plans, meeting notes. Learn the tool's edges on low-stakes work before the high-stakes stuff.
You still own the output
Verify facts, numbers and names. The judgment is yours; the typing is what you delegated.
The den · Featured
Latest from the bear
The longer versions of my LinkedIn posts — one real moment, one thing worth stealing. Tap any card to read the whole thing.
Insight 03 · From the field
6 min readAI took 20 minutes off my grocery run. Then it tried to age my tomatoes.
Where a everyday task turns into a lesson about first drafts, and why the second ask is almost always the better one.
#theaverageaibear#aiforbeginnersRead
Insight 05 · At work
5 min readOnboard AI like it's the new hire named Sam
You wouldn't drop a one-line task on a new colleague's desk and walk away. Same rules, same probation period.
#theaverageaibear#aiRead
Insight 06 · Most read
4 min read"Your dad would be proud." Thanks, AI — but that's not feedback.
Placation is a feature. Here's the one line that turns cheerleading into a genuinely useful critique.
#theaverageaibear#aiforbeginnersRead
Insight 07 · Off the clock
7 min readI let AI ride shotgun on a trip before I trusted it with anything that mattered
Why low-stakes practice is the fastest, least painful way to learn where a tool over-promises.
#theaverageaibear#practiceRead
The archive
Earlier from the den
You lost me at "cloud"
Four main AI tools exist. You don't need to master them — pick one and try.
AI is more than a fancy search engine
Stop searching, start asking. How to tell it what you actually want.
Every spring I hit the same wall
Camps, deadlines, logistics — and how I handed the planning over.
Using AI to fix my AI fatigue
My "fabulous 15" setup: how agentic tools build skills, not just spit out answers.
"Is that from me, or did you infer it?"
Catching the quiet assumptions — and the one prompt that surfaces them.
A partner in reasoning
What it's like to watch your own thinking become visible, and check your gut against it.
Weekly bear tips
Small things that make a real difference
One-minute reads: a prompt idea, a productivity trick, or a gotcha I hit so you don't have to. Refreshed as I learn them.
Give it the bad example too
Show one output you liked and one you didn't. Contrast teaches faster than adjectives.
It will invent a citation
If a link, stat or policy number matters, open the source yourself. Every time.
End with "what did you assume?"
A two-second question that exposes the guesses baked into the answer.
Paste your own writing
Three paragraphs of your voice beats any "write in a friendly tone" instruction.
Long chats drift
When answers get mushy, start fresh and re-paste only what still matters.
Ask for the plan first
"Outline your approach before writing" catches wrong turns while they're cheap.
Prompting corner
Better in, better out
Prompting isn't a secret language. It's briefing someone capable who knows nothing about your world.
Prompt of the week
"Act as the person who has to live with this decision for a year. What do you wish we'd thought about today?"
Great for plans, policies and anything you're about to send to a room full of people.
Skip the encouragement. Be direct: where is this weakest, and what would a skeptic push back on first?
Here's the goal, the audience, and one example of good. Ask me three questions before you start.
List every assumption you made that I didn't tell you. Mark the ones you're least sure about.
Explain this to a smart colleague who doesn't work in tech. No acronyms, no metaphors about rockets.
Bear necessities · Resources
Things I'd point a friend to
Books, courses, newsletters and tools that earned their place. No affiliate anything — just what I've actually found useful.
One Useful Thing — Ethan Mollick
A professor who tests AI on real work and writes plainly about what helps. Warm, practical, no hype.
AI for Everyone — Andrew Ng
Short, friendly and non-technical, from one of the field's founders. Zero math. Made for the rest of us.
Hard Fork
Weekly AI and tech news that's actually a good time. Perfect for the commute or the dishes.
Anthropic's blog
What one of the big AI labs is building, in their own words. Sometimes dense, but straight from the source.
Microsoft Copilot
If your workplace already runs on Microsoft 365, this is the lowest-friction place to start experimenting.
Co-Intelligence — Ethan Mollick
The book I hand to people who want one thing to read before forming an opinion.
Down in the den
Longer thoughts, half-finished experiments, and the odd rabbit hole
Not every idea fits in a post. This is where the slower writing lives — the reflections, the experiments that failed interestingly, and the questions I'm still chewing on.
What changes when the tool argues back
Three weeks of asking AI to disagree with me before agreeing.
The citizen developer's quiet advantage
Why not knowing how it works can make you a better user of it.
Notes from a very average experiment
An honest log of what broke, what surprised me, and what I'd redo.
The sweet spot
Where AI genuinely improves an ordinary work week
Not the moonshots. The everyday places where it earns its keep for people who don't write code.
01
Turning a messy meeting into decisions
Paste the notes, ask for decisions, owners and open questions. Ten minutes back, every time.
02
Writing the first version of anything
Procedures, emails, briefing notes. The blank page problem, mostly solved.
03
Translating between audiences
The same update for a technician, a manager and an executive — without three drafts.
04
Interrogating a long document
Ask what's missing, what contradicts, what a reviewer will flag first.
05
Learning a tool while using it
"Explain what you just did and why" turns every task into a small lesson.
06
Planning the annoying stuff
Schedules, logistics, checklists. Low stakes, high relief.
The bees buzz
What's making noise this month
A curated, jargon-free look at the AI news actually worth your attention. Placeholder cards for now — this fills up as things happen.
Agentic tools go mainstream
What it means when the assistant starts taking multi-step action on your behalf.
// coming soon
AI inside the tools you already use
The quiet arrival of assistants in the software your workplace already pays for.
// coming soon
Policy catches up, slowly
Workplace AI guidelines are landing. What to read before you paste anything sensitive.
// coming soon
Fun corner
A little something for the slow afternoons
There's a small on-theme time-killer coming to this spot — a proper excuse to procrastinate in good company. Consider this the honey pot being set out.
Mini-game · in the works
FAQ
Questions that come up a lot
Short answers, no hedging.
Ask the bear
Questions, comments, hellos
Wondering something everyone else seems to already get? Ask it here. No question is too small or too basic — if you're wondering it, someone else is too. I read every one, and the good ones become posts.