Getting real
work out of AI

The tools all do roughly the same thing: you type what you want in plain English, they write back. You don't need to learn a new language or any special syntax. If you can describe the task to a sharp new hire, you can do it here.

There are four general-purpose ones worth knowing. Pick one and get comfortable — they're more alike than different. Each has a free tier, plus a paid plan (around $20/month) that unlocks the smartest models and the image tools — and gives you a lot more usage time.

Picking a model — and what it costs you

Inside a tool you'll often get to choose which model answers — a fast everyday one and a more powerful one. (On Claude it's Sonnet for everyday work and Opus for heavy lifting; ChatGPT and Gemini have the same idea under different names.) The more powerful model can reason harder and hold more in its head at once — but it costs more to run and burns through your usage allowance faster.

The rule of thumb: let the standard model handle the everyday work, and only switch up to the powerful one when the task genuinely calls for deep thinking — dense analysis, a long and complicated document, anything where it needs to hold a lot together and reason carefully. For quick drafts, rewrites, and summaries, the standard model is plenty — and staying on it keeps you from burning through your usage limit halfway through the day.

The single skill that separates a frustrating answer from a useful one isn't technical. It's how clearly you ask.

A vague request gets a vague, generic answer. The fix is to tell it three things: who it should be, what you want, and what "good" looks like. Watch the difference.

The four moves that fix almost everything

The tool only knows what you tell it. It can't see your files, your past orders, or how your company does things — unless you put that in front of it. The more relevant material you give it, the more the answer sounds like your business instead of a generic one.

One caveat: tools wired into your accounts — like Copilot inside Microsoft 365 — can already reach some of your files and email on their own. Here we're mostly talking about the public chatbots (ChatGPT, Claude, Gemini), which know nothing about your business until you hand it to them.

Ways to hand it context

If you find yourself typing the same setup over and over, you can save it once and reuse it forever.

Every major tool lets you save a reusable setup like this: write the instructions and load the background once, give it a name, and from then on it knows the role and the context — you just hand it each new task. But the names hide a real distinction. ChatGPT's version is a Custom GPT; both ChatGPT and Claude also have Projects; Gemini has Gems. A Custom GPT and a Project are not the same thing — they're built for two different jobs, and it's worth knowing which one you want.

Custom GPT vs. Project — same engine, different job

Both save the same two things: your instructions and your background material. The difference is what they're built to do with it — one is for handing out, the other is for working in.

(Gemini's Gems are a third flavor — free to build and use — but they live in your own Gemini rather than being handed around.)

What this means for sharing

Before you promise something "for everyone," match the tool to the goal. If the point is to hand it around, build a Custom GPT — a link works even from one person's paid account. If the point is to organize your own ongoing work, use a Project. A genuinely shared team workspace — one Project everyone edits — is a separate, paid step, worth taking once it's proven itself.

Good candidates to build: anything you do repeatedly with a consistent setup — drafting replies in your company's voice, turning purchase orders into a clean checklist, writing product descriptions, proofreading to a house style. Start one for yourself; grow it into a shared tool when you know it earns its keep.

You don't need a big project to get value today. These are the small, daily things that quietly give you back time — start with whichever one you'd use this week.

Two very different things get lumped together as "AI images." Knowing which one you need saves a lot of wasted time.

Generating vs. editing

The common, useful jobs

Putting your product in a real scene

The most useful move for a product business: start with a plain product shot, add branding, and drop it into a lifestyle setting — no photoshoot required. Three steps:

The standout tool for this is Google's Gemini — its image engine (nicknamed "Nano Banana") is unusually good at keeping a real product looking like itself while changing everything around it. ChatGPT can do it too. You hand it the blank, the logo, and a sentence describing the scene. It's not flawless — AI can nudge small details, so this is for concepting and listings, not final production art.

A few simple habits keep you on the right side of this — using it confidently without getting burned.

You'll hear "agentic AI" thrown around constantly. Here's what it actually means — and what it can and can't do for a business your size.

Everything so far has been AI that answers. Agentic AI is the next step: AI that acts.

What "agentic" actually means

An agent is AI you hand a goal to instead of a single question. It plans the steps, uses tools to carry them out, and works through the task — checking its own progress as it goes. The shift is from it tells you to it does it.

You'll see this dressed up a dozen ways — "agents," "agent mode," "computer use," "Operator," "autonomous AI." Under the marketing, it's all the same idea: AI that takes actions, not just AI that talks.

What you can do with it yourself

At a personal or small-business level, agentic AI is mostly about connecting your apps and automating multi-step chores — and a lot of it needs no developer. A few real examples:

The no-code tools that make this possible: Zapier, Make, and n8n for wiring apps together; Power Automate / Copilot Studio if you're a Microsoft shop; and the "agent" modes now built into ChatGPT, Claude, and Gemini themselves. You point and click your way to a working automation.

Where it gets bigger — and where we come in

The moment an automation has to touch your real systems, handle sensitive data, run reliably every single time, or stitch several pieces of software together at scale, it stops being a no-code weekend project and becomes real engineering. That's the line where you want the big guns instead of wrestling it yourself — and that's exactly what Rogue Agents does: purpose-built AI systems wired into how your business actually runs.

All of this was high-level on purpose. I'm not trying to teach you everything about AI — this is a quick, referenceable cheat sheet: what's what, the lay of the land, something to come back to as you keep exploring on your own.

For the rest of our time, let's get practical. I want to hear what you're actually working on day to day, and how you're using AI on it right now. Throw out any questions as we go — I'll help you do what you're already doing, and show you tips and tricks to do it better.

We'll do it live, together, so everyone picks up the little nuggets they didn't know. That's the part that sticks.