What's the first thing I should try with AI if I've never used it before?

The first thing you should try with AI is summarizing something you already understand well, because that lets you immediately judge the output's accuracy without needing to fact-check every sentence. Most beginners make the mistake of asking AI about topics they know nothing about, which leaves them unable to tell if the response is brilliant or confidently wrong. Start with something familiar. Copy-paste a long email thread you've been part of and ask the AI to summarize the key decisions in three bullet points. You'll see instantly whether it caught the nuance or missed the point. This builds the most important AI skill: calibrated trust. You learn when the tool is reliable and when it's just guessing. I've seen people try AI for the first time by asking it to write their resume or explain quantum physics, then get frustrated when the output needs heavy editing. That's like learning to drive on a Formula 1 track. Another good starter task: ask the AI to rewrite a paragraph you've already written in a different tone — make it more casual, more professional, or shorter. You know what you meant, so you can evaluate whether the rewrite preserved your intent. This teaches you how to give direction, which is the core skill behind prompt engineering. For a deeper dive, see our guide on how to write AI prompts. A specific example: take a work email you sent last week that felt too long. Paste it in and say 'Rewrite this in three sentences while keeping the friendly tone.' The result will either nail it or miss the mark, and either way, you've learned something about how the AI interprets tone. The goal isn't to get perfect output on day one. It's to develop an intuition for what the AI is good at and where it stumbles. **Related**: How do I know if AI is making things up? | What's the difference between ChatGPT and an AI writing tool?