AI is not magic. It is more like a very fast, very well-read assistant who never gets tired. Some tasks it handles brilliantly. Others it fumbles. Knowing the difference saves you a lot of frustration.
Writing — the first draft you never have to dread again
Staring at a blank page is the worst part of writing. AI fixes that. You give it a rough idea and it hands you a starting point you can edit.
Say you need to write a thank-you email after a job interview. You can tell AI: "Write a short thank-you email to a hiring manager after a software developer interview. Keep it warm but professional." You get a solid draft in seconds.
You still read it, tweak it, make it sound like you. But the blank page is gone. That alone saves enormous mental energy every single day.
Think of AI as your first-draft machine. You are still the editor. Your voice, your judgment — that is what makes the final version actually good.
Summarising — turning long things short
Got a 10-page report you need to understand in 5 minutes? Paste it into an AI chatbot and say: "Summarise this in plain English. Three bullet points max." Done.
This works for articles, meeting notes, long email threads, even boring legal documents. Tina Huang, an AI educator on YouTube, calls this one of the most immediately useful skills anyone can learn.
One warning: always skim the summary yourself. AI occasionally misses something important or smooths over a detail that actually matters.
Paste a long article into your chatbot right now and type: "Give me the 3 key takeaways in plain English." That is your first real AI win.
Explaining — learning anything, at your own pace
AI is a patient teacher that never sighs when you ask the same question twice. You can ask it to explain compound interest, how a vaccine works, or why your lease agreement has that weird clause.
The trick is to tell it your level. Say: "Explain this like I am 12 years old" or "Explain this to someone who has never studied economics." The explanation you get back will actually make sense.
Safe to ignore for now: tools like NotebookLM take this further — you upload a document and it becomes a kind of interactive tutor for that specific content. But even a basic chatbot handles most explaining jobs well.
Brainstorming — getting unstuck in minutes
Brainstorming with AI feels less like asking a computer and more like talking to someone who has read everything. You throw out a half-formed idea and it throws back ten angles you had not considered.
Say you are planning a team offsite and have no ideas. Ask: "Give me 10 low-cost team offsite ideas for an 8-person remote company that likes the outdoors." You will have more ideas than you need in under 30 seconds.
You do not have to use any of the suggestions directly. The point is to break the mental block. Domain knowledge matters here too — you can spot the bad ideas because you know your situation. AI does not.
Safe to ignore for now: AI can also brainstorm code, marketing strategies, and business plans. Those uses come later. For now, just practise on low-stakes everyday problems.
The honest truth about what AI cannot do
AI does not know your life, your relationships, or what actually happened in that meeting. It cannot check if something is true right now. And it sometimes sounds very confident while being completely wrong.
Tina Huang puts it bluntly: the people who get the best results are the ones who know what a good output looks like. That means you need to bring your own judgment every single time.
AI is a tool that amplifies what you already know. If you bring curiosity and a critical eye, it multiplies your output. If you copy everything blindly, the results will show it.
Here's something I need to understand quickly: [paste an email, article, report, or meeting notes] Summarise it for me: — The main point in one sentence — 3 things I need to know — Any action required (if any)