Six months ago, my mother called me and said, “Everyone at book club is talking about ChatGPT. What is it? Should I be worried?”
I gave her my elevator pitch: “It’s a computer program you can talk to. You ask it questions, and it answers them. You can ask it to write things, explain things, or help you think through problems. Think of it as a really smart assistant that knows about almost everything but occasionally makes stuff up.”
She tried it that evening. The next morning she texted me: “I asked it to help me write a letter to the insurance company about that denied claim. It was better than anything I could have written. This changes everything.”
She wasn’t wrong.
Start Here, Not Everywhere
The biggest mistake I see people make: trying to learn “AI” as a subject. Don’t do that. Pick one tool, use it for one thing, and get good at that one thing. Then expand.
If you’ve never used AI: go to chat.openai.com (ChatGPT) or claude.ai. Don’t create an account yet — just try it. Type a question you’d normally Google. Something specific: “Explain how a 401k match works, using simple terms, for someone who just started their first job.”
Read the response. Notice that it’s not a list of links — it’s a direct, clear explanation written for your specific question. That’s the fundamental difference between AI and search. You ask a question. You get an answer.
If you’ve tried it but aren’t using it regularly: you probably haven’t found your “aha” use case yet. Here are the ones that convert skeptics:
Write a difficult email. The kind you’ve been putting off because you can’t find the right words. Describe the situation to ChatGPT: “I need to email my landlord about recurring plumbing issues. I’ve asked twice before and nothing happened. I want to be firm but not hostile. The issues are [list].” The draft you’ll get back is better than what most people write after agonizing for an hour.
Summarize something long. Paste in an article, a report, or a document you don’t want to read. “Summarize the key points of this document in 5 bullet points.” I’ve saved hundreds of hours doing this with reports, research papers, and legal documents.
Learn something new. “Explain blockchain to me like I’m 12.” “What’s the difference between a Roth IRA and a traditional IRA, and which is better for someone earning $75K/year?” The explanations are patient, clear, and customized to your level.
The Tools You Actually Need
ChatGPT is the starting point for most people, and that’s fine. The free tier is genuinely useful. It handles writing, analysis, brainstorming, coding, and conversation. If you’re going to pay for one AI tool, ChatGPT Plus at $20/month gives you GPT-4o and image generation.
Claude is what I use most. I find Claude’s responses more thoughtful and nuanced, especially for complex analysis, long-form writing, and ethical questions. It handles long documents well — paste in a 50-page report and ask questions about it. Free tier available, Pro at $20/month.
Perplexity is the AI search engine. For any research question — “What are the current mortgage rates for a 30-year fixed in California?” or “What are the side effects of metformin?” — Perplexity gives you a direct answer with cited sources. I’ve largely replaced Google with Perplexity for research. Free tier available.
GitHub Copilot is for people who write code (or want to start). It suggests code as you type, explains existing code, and helps debug problems. Even if you’re learning to code, Copilot accelerates the process enormously. $10/month.
What AI Is Good At (And What It’s Not)
Good at:
– Writing first drafts of anything (emails, reports, posts, letters)
– Explaining complex topics in simple terms
– Summarizing long documents
– Brainstorming ideas
– Answering factual questions (usually)
– Basic data analysis
– Writing and debugging code
– Translation
Not good at:
– Being right 100% of the time (it will confidently state incorrect things)
– Understanding your personal situation deeply
– Replacing professional advice (legal, medical, financial)
– Creating truly original creative work (it recombines, it doesn’t innovate)
– Knowing what happened yesterday (training data has a cutoff)
– Keeping secrets (don’t share sensitive info you wouldn’t share with a stranger)
The Single Most Important Skill
Learning to talk to AI effectively — prompting — is the single most valuable skill in the AI era. The difference between a bad prompt and a good one is the difference between a useless response and a life-changing one.
Bad: “Help me with my resume.”
Good: “I’m a marketing manager with 8 years of experience applying for a VP of Marketing role at a B2B SaaS company. Review my resume and suggest improvements that highlight leadership experience and data-driven decision making. Here’s my resume: [paste]”
The good prompt includes who you are, what you’re trying to achieve, what kind of output you want, and the relevant context. This isn’t complicated — it’s just being specific.
Other tips that make a big difference:
Ask for a specific format. “Give me 5 bullet points” or “write a 200-word paragraph” or “create a table comparing options.”
Tell the AI who it’s writing for. “Explain this to a 10-year-old” vs. “explain this to a software engineer” produces very different (and appropriate) responses.
Iterate. If the first response isn’t quite right, say “this is close but make it more concise” or “good, but emphasize the cost savings more.” AI conversations are conversations — you refine as you go.
Staying Safe
Don’t trust it blindly. AI makes mistakes. Verify important facts. Check numbers. Confirm that the legal advice isn’t outdated or jurisdiction-specific. “Trust but verify” is the right mindset.
Don’t share sensitive data. Unless you’re using an enterprise version with data privacy guarantees, assume that anything you type could be stored and potentially used. Don’t paste confidential business documents, personal health records, or financial information.
Be transparent about AI use. If you’re submitting AI-assisted work, know the rules. Some schools prohibit it. Some employers welcome it. Some clients don’t care. Know the expectations before you submit.
What Happens Next
AI is going to get better. Fast. The ChatGPT you use today will look primitive compared to what exists in two years. Prices will drop. Capabilities will increase. Integration with your existing tools will become smooth.
The people who start using AI now — even imperfectly, even for small things — will have a massive advantage over those who wait. Not because AI is magic, but because it’s a skill. And like any skill, the people who practice earliest get the furthest ahead.
Open a tab. Go to ChatGPT or Claude. Ask it something. Start there. Everything else follows.
🕒 Last updated: · Originally published: March 15, 2026