I keep meeting smart people who feel like AI passed them by. They open ChatGPT, type one question, get a weird answer, and quietly close the tab. Then they see someone on LinkedIn talking about “agents” and “APIs” and decide the whole thing isn’t for them.
So when I stumbled on this post from a self-described non-technical LinkedIn creator, I had to share it. The author doesn’t write code. Doesn’t pay for every AI tool. Doesn’t chase every launch. And still uses AI daily to do things that were out of reach a year ago. That’s the story I think more beginners need to hear.
🧭 The author’s honest list of “I don’t”
The post opens with a list that reads like a confession. I found it refreshing. This is what the original poster admits to:
- Not writing code
- Not paying for all the AI tools
- Not jumping on every AI launch
- Asking Claude to explain things twice
- Describing what they want in plain English
- Telling ChatGPT when they don’t understand
- Writing “explain it more simply” right in the prompt
- Asking for instructions one step at a time
- Learning technical AI terms only when needed
- Sharing screenshots when words won’t do
Quick translation for anyone brand new: a “prompt” is just the message you type to an AI. Claude and ChatGPT are two of the most popular AI chat assistants. That’s it. Nothing on that list requires a technical background. It requires being honest about what you don’t know and saying so out loud.
🚫 The scary words you can skip for now
This industry pro names the exact list that overwhelms most beginners: Python, APIs, Agents, MCPs, Automations. If your stomach dropped reading those, you’re not alone. Let me explain each in one line so they lose their power:
- Python: a programming language. You don’t need it to chat with an AI.
- APIs: a way for two programs to talk to each other. Developers use them. You probably won’t.
- Agents: AI that can take actions on its own, like browsing a site or filling a form. Cool, but advanced.
- MCPs: a plug-in standard that lets an AI connect to other tools. Again, advanced.
- Automations: setting things up to run without you. Useful later, not on day one.
The author’s point is simple. You don’t need to learn any of this before you get real value. The list is for people building AI products, not for people who want to get through their Tuesday faster.
🎯 Start with the job, not the tool
This is the part I think is the real gold. Before you learn anything, the original poster says to ask yourself one question: what do you actually want to do in your work? The post gives five examples that most of us will recognize:
- Build a small website
- Understand a spreadsheet
- Record a meeting without bots
- Type faster to prompt and work faster
- Finally get through a 40-page document
Pick one. Just one. Then find the shortest path to doing it with AI. If your problem is the 40-page document, you don’t need a course. You paste the text into Claude or ChatGPT and type “summarize this in plain language and list the five things I need to act on.” Done. You just used AI.
The author compares it to a bike. You don’t study physics before riding. You get on, wobble, and about 23 minutes later you’re pedaling by yourself. The same goes for AI. You need a proper start, not a full education.
🙅 What this creator refuses to teach beginners
I loved this bit. The post’s author lays out a list of things they deliberately won’t cover for people who are just starting:
- NLP (natural language processing, the field behind how AI reads text)
- Agent harnesses (the scaffolding developers build around AI agents)
- AI doomerism or overhype
- What an MCP is
- How an attention matrix works (the math inside the model)
- Sequencing, embeddings, or LLMs (LLM means “large language model,” the engine behind Claude and ChatGPT)
Notice a pattern? Every one of those is about how AI works on the inside. None of it helps you finish a spreadsheet by Friday. The author’s stance is that you’re allowed to need help and you’re allowed to ask stupid questions. That mindset is a pretty good place to begin.
Why it matters: beginners don’t quit because AI is hard. They quit because the vocabulary around it makes them feel behind before they even start. Strip the vocabulary and most of the barrier disappears.
🛠️ The five things worth learning first
The creator boils the essential starter kit down to five items. I’d hand this list to anyone opening ChatGPT for the first time:
- Which AI to use (pick one assistant and stick with it for a month)
- How to prompt it (plain English, one request at a time)
- The few features you’ll actually use most (file uploads, voice input, and “explain it simpler”)
- One practical example of being productive with AI, start to finish
- Reusable “skills,” which are saved instruction sets you can drop into any AI so you never retype the same setup
💡 Tips you can use today
- When an answer confuses you, reply with “explain it more simply.” It works every time.
- Ask for one step at a time. Say “give me step 1 only, then wait for me.”
- Stuck describing something? Take a screenshot and upload it. Both Claude and ChatGPT can read images.
- Learn a technical word only when it blocks you. Not before.
- Don’t buy every tool. One free chat assistant covers most beginner jobs.
Prompt of the Day, in the spirit of this post: “I’m a beginner. I want to [your task]. Explain what to do one step at a time, in plain English, and check that I understand before moving on.”
If you’ve been waiting for permission to start using AI without learning to code, this is it! Head over to the full LinkedIn post and read the author’s whole list. It’s short, honest, and might be exactly the nudge you needed.