What can AI actually do for me?

by Stephen Wu, a retired engineering manager, spent several years building AI workflows for his teams and now trying to make AI something everyone can understand and use successfully

AI tools like ChatGPT, Claude, and others are conversation partners that help you think through problems, understand information, and get things done. Instead of searching through articles or struggling alone, you describe what you need and AI works with you to figure it out.

The key is understanding what AI excels at—and what it struggles with—so you know when to use it and when to handle something yourself.

The Steps: Understanding AI's capabilities

What tasks is AI good at vs. not good at? - Learn what AI excels at (processing information, generating variations, explaining concepts) and where it falls short (current events, specialized expertise, fact verification).

How do I know when to use AI vs. do it myself? - Develop a practical decision framework for when AI will actually save you time versus when doing it yourself is faster.

What's the difference between ChatGPT, Claude, and other AI tools? - Understand how the major AI chat tools compare and whether specialized tools like Perplexity or Grammarly offer advantages for specific tasks.

Do I need to pay for AI or are free versions enough? - Figure out if free AI tools meet your needs or if paid versions ($20/month typically) offer meaningful benefits for your usage patterns.

A Real Example: Planning a career transition

Let's walk through how someone might use AI to explore changing careers from marketing to UX design. This example shows AI's capabilities in action—and its limitations.

Sarah works in marketing but is interested in UX design. She's not sure if it's a realistic transition or where to even start. She opens an AI chat.

AI prompt: "I work in digital marketing and I'm interested in transitioning to UX design. I don't have a design background. Is this a realistic career change and what would I need to learn?"

AI responds with an overview: yes, marketing to UX is a common transition because both involve understanding user needs. AI outlines core UX skills (user research, wireframing, prototyping, usability testing) and notes that Sarah's marketing experience with customer insights is relevant.

This is AI doing what it does well—processing general knowledge about career paths and synthesizing an overview quickly.

AI prompt: "What specific skills from my marketing background would transfer to UX design? I've done customer surveys, A/B testing for email campaigns, and created user personas for targeting."

AI explains how each skill maps to UX work. Customer surveys are similar to user research. A/B testing translates to usability testing. User personas in marketing are nearly identical to personas in UX design. AI helps Sarah see connections she might not have recognized immediately.

She wants to understand the learning path.

AI prompt: "What do I need to learn to become job-ready in UX design? Give me a structured learning path starting from my marketing background."

AI provides a multi-phase plan: start with UX fundamentals, learn design tools like Figma, practice creating wireframes and prototypes, build a portfolio with redesign projects, study interaction design principles, and learn basic user research methods.

This is useful for brainstorming and structure, but Sarah knows she needs to verify this with actual UX designers and job postings. AI gives her a starting framework, not definitive career advice.

AI prompt: "Can you suggest 5 portfolio projects I could do to demonstrate UX skills to potential employers, given that I don't have client work yet?"

AI suggests specific project ideas: redesign a frequently-used app's checkout flow, create a case study analyzing a website's usability issues and proposing solutions, design a mobile app for a common problem, conduct user interviews about a product and create personas, and document a comparative analysis of competitor products.

These are concrete starting points. Sarah can take any of these ideas and execute them, documenting her process for a portfolio.

But when Sarah asks about salary expectations for junior UX designers in Austin, Texas in 2025, AI's response is less reliable. It might offer outdated figures or national averages that don't reflect Austin's current market. For current, location-specific salary data, Sarah needs to check job boards, Glassdoor, or talk to people in the field.

AI prompt: "What are the typical day-to-day responsibilities of a UX designer at a tech company versus an agency? I want to know what the actual work feels like."

AI describes typical workflows, meetings, deliverables, and challenges in both environments. But this is generalized information. To really understand what the work feels like, Sarah needs to talk to actual UX designers—something she might ask AI to help her prepare for.

AI prompt: "I want to reach out to UX designers on LinkedIn to ask about their career paths. Can you help me write a message that's professional but not overly formal? I want to ask if they'd be willing to do a 20-minute informational interview."

AI drafts a message. Sarah reads it, adjusts the tone to sound more like herself, and uses it as a template for outreach.

Throughout this process, AI helps Sarah:

Quickly understand if a career change is feasible
Map her existing skills to a new field
Generate a structured learning plan
Brainstorm portfolio projects
Draft outreach messages

But AI doesn't replace:

Current salary data
Actual conversations with UX professionals
Hands-on practice with design tools
Real feedback on her work
Verification of job market realities

Sarah uses AI to process information, generate options, and think through decisions faster than she could alone. Then she takes AI's outputs and tests them against reality—checking job postings, talking to designers, and practicing actual UX work.

Tools: Where to start

You don't need to download anything or create special accounts to start. If you're reading this, you likely already have access to AI through a web browser.

Open ChatGPT, Claude, Microsoft Copilot, or Google Gemini—all have free versions. Type a question or describe a task. See what happens. That's it.

AI prompt: "I'm brand new to using AI. What are some simple tasks I could try right now to see how AI can help me in my daily life?"

AI will suggest practical starting points based on common needs. Try a few, see what's useful, and ignore what isn't.

If you find yourself using AI frequently and hitting message limits on free versions, you might consider paid plans. But most people find free versions completely sufficient for months or years.

Important Reminders

AI is a thinking partner, not an oracle. It helps you process information and generate options faster than you could alone, but it doesn't replace your judgment, verification, or real-world action.

When AI gives you information, especially about current events, facts, or specialized topics, verify it. Check sources, test the advice, ask experts. AI is excellent at synthesis and brainstorming but not at guaranteeing accuracy.

You'll learn what AI is good at through use. Some tasks that seem perfect for AI turn out to be faster to do yourself. Other tasks you wouldn't expect AI to help with turn out to be incredibly useful. Experiment with real tasks rather than hypothetical ones.

Start simple. Ask AI a genuine question you have right now. See if the response helps. Refine your question based on what you get back. You'll develop intuition for how to work with AI through practice, not by reading about it.