AI Basics

Everything You Need to Know
to Actually Use AI Tools

No jargon. No hype. Just the practical fundamentals — what these tools do, how to talk to them, and how to build your first real workflow.

Get the Free Starter Toolkit

5 Things You Need to Understand First

Before you touch any AI tool, these are the five concepts that make everything else click.

01

AI tools are text machines — not minds

ChatGPT doesn't understand your work the way a colleague does. It predicts useful text based on your input. The better your input, the better the output. There's no magic — only patterns. Understanding this removes unrealistic expectations and makes you a better user.

02

The prompt is the skill

Your results depend almost entirely on how you ask. A vague prompt gets a generic answer. A specific prompt with context, format instructions, and a defined output gets something usable. Prompting is the skill you're actually learning.

03

AI doesn't replace your judgment — it removes the blank page

AI is fastest at first drafts, research summaries, and repetitive writing. The harder your task, the more your judgment matters on top of the AI output. Think of it as a very fast first-drafter — not an expert making decisions for you.

04

Each tool has a job

ChatGPT writes and thinks. Notion organises and stores. Canva designs. Beehiiv sends email. Make automates. Trying to use one tool for everything leads to frustration. The DSHQ AI Systems are built around this: the right tool in the right role.

05

Start narrow, then expand

The biggest beginner mistake is trying to automate everything at once. Pick one repetitive task you do weekly — email drafting, social posts, research — and make AI useful there first. Build the habit before building the system.

The Anatomy of a Good Prompt

Most bad AI results come from prompts that are too short and too vague. Here's the structure that works.

Role:     You are a professional email writer for solo business owners.

Context:  I run a content writing service. My client, a marketing agency,
          hasn't paid their invoice for 3 weeks despite two reminders.

Task:     Write a firm but professional follow-up email that maintains
          the relationship while making clear this needs to be resolved
          this week.

Format:  Short — under 150 words. Professional tone. No passive language.

Role

Tell it what kind of expert to act as. This sets the voice and expertise level of the response.

Context

Give it the specific situation. The more relevant detail you provide, the more targeted the output.

Task + Format

State exactly what you want and how you want it — length, tone, structure, output type.

The free toolkit includes 50 prompts built on this structure — tested and ready to use. Download them →

Build Your First AI Workflow in 4 Steps

A workflow is just a repeatable process. Here's how to build your first one — start to finish.

1

Pick one repetitive task

Something you do weekly that involves writing, research, or summarisation. Good first picks: drafting emails, writing social captions, summarising meeting notes, or generating content ideas.

2

Write a prompt for it using the Role-Context-Task-Format structure

Don't worry about perfection. Start with something specific and iterate. The prompt doesn't have to be long — it has to be clear.

3

Run it 3–5 times and refine

Each run teaches you something. Add more context. Adjust the tone instruction. Specify the output length more clearly. This iteration loop is where the skill is built.

4

Save the working prompt in Notion

Once a prompt works well, save it as a reusable template. This is the beginning of your personal prompt library — and the foundation of every DSHQ system.

Take It Further

Free Resource

Get the 50-Prompt Starter Pack
and the AI Tool Stack Guide

Everything in the free toolkit is built around these basics — ready to use the moment you open them.

Get Free Toolkit