How to Use AI to Learn New Software Without Watching Hours of Tutorials

I stared at Blender's interface for the third time this year, closed it, and went back to pretending I didn't need to learn 3D modeling. That's been my pattern with complex software for years. Download it, get overwhelmed by the 47 toolbars, promise myself I'll "find a good tutorial series," never do it. Then about three months ago I accidentally discovered that ChatGPT could teach me software in a completely different way, and now I've actually learned four tools I'd been avoiding.

The Method I Stumbled Into

It started because I was too lazy to search YouTube. I had DaVinci Resolve open, wanted to do one specific thing — add a simple zoom effect to a clip — and just asked ChatGPT how to do it. Not "teach me video editing." Just "I have a clip selected in DaVinci Resolve, how do I make it slowly zoom in over 3 seconds."

The answer was step-by-step. Click here. Look for this panel. If you don't see it, go to Window and enable it. I followed along with the software open next to the chat. Took maybe 90 seconds. Done.

Then I needed to know how to add a crossfade. Another question. Another 90 seconds. By the end of an hour, I'd edited an entire short video by just asking questions as they came up. No 40-minute tutorial about the history of nonlinear editing. No waiting for the YouTuber to get to the part I actually needed.

This felt different. I was learning by doing, but with an infinitely patient guide who only answered what I asked.

What Actually Works (and What Doesn't)

I've now tried this approach with Blender, Notion, Figma, and Obsidian. Here's what I've figured out.

Be stupidly specific about where you are. "How do I crop an image in Figma" gives you generic steps. "I have an image selected on my canvas in Figma and I want to crop it to show only the top half" gets you the exact keyboard shortcut and the specific behavior to expect. The AI can't see your screen, so you have to be its eyes.

Screenshots help enormously if you're using GPT-4. I've uploaded images of an interface I didn't understand and asked "what is this panel for and do I need it right now." Sometimes the answer is "you can ignore that entirely for what you're doing." That permission to ignore things is worth a lot when you're overwhelmed.

Here's what doesn't work: asking broad questions. "Teach me Blender" produces the same generic overview you'd get anywhere. The magic happens when you're in the middle of trying to do something and you hit a wall. That's when AI becomes useful — not as a course, but as a coworker who happens to know the software.

Also doesn't work: trusting it completely. About 15% of the time, the instructions are slightly wrong. A button moved in a recent update. A shortcut that works on Mac doesn't work on Windows. The menu is named something different in my version. This burned me twice in Blender before I learned to double-check anything that feels off.

The Kick: The "Why Does This Exist" Question

Here's the thing nobody mentions about learning software this way. You can ask questions that tutorials never answer because they assume you don't need to know.

I spent ten minutes in Notion completely confused about why there were both "databases" and "linked databases" and when I'd use one versus the other. Every tutorial I'd seen just showed you how to make them, not why you'd choose between them. So I asked ChatGPT: "Explain like I'm frustrated and confused why Notion has both databases and linked views of databases. When would I actually need the linked version?"

The answer finally made it click. And then I asked follow-ups. "So if I delete the linked view, the original data stays?" Yes. "What if I filter the linked view, does that affect the original?" No. Five minutes of Q&A gave me more understanding than the 20-minute "Notion databases explained" video I'd watched twice and retained nothing from.

This works for every piece of confusing software design. Why are there three different ways to do this? Why does this option exist? What happens if I mess this up? These are the questions you're embarrassed to Google but will absolutely ask an AI that doesn't judge you.

I now start every software learning session by identifying the thing that confuses me most and just asking about it directly. The friction that used to make me quit is now the first thing I address.

The Real Limitation Nobody Talks About

There's a ceiling to this. AI can tell you how to do isolated tasks, but it can't tell you what tasks you should be doing. It doesn't know your project. It can't say "actually, you're overcomplicating this, just use a different approach entirely."

I learned this the hard way in Blender. I spent 45 minutes asking how to model something using technique A, when technique B would have taken five minutes. The AI answered every question correctly. It just couldn't see that I was on the wrong path because I didn't know enough to ask the right question.

For this, you still need a human. A forum. A subreddit. A friend who actually uses the software. Someone who can look at what you're trying to accomplish and say "why are you doing it that way?"

But for the 80% of learning that's just "how do I do this specific thing I already know I need to do," the AI approach has genuinely changed how I approach new tools. I downloaded Affinity Photo last week and felt zero dread about learning it. I'll just ask as I go.

The weird part is I'm now suspicious of traditional tutorials. All that time spent watching someone else do things, hoping I'll remember when I need it. When I could just... ask. Though I still wonder if I'm missing something by not learning the "proper" foundations. Maybe there's a reason everyone

Heads up: Some links in this post may be affiliate links. I only recommend tools I've personally tested. Opinions are entirely my own.