Why I Started Teaching AI to Regular People

Okay so here's the thing. I didn't start this because I had some grand plan to educate the masses or whatever. I started because my mom called me crying, not literally crying, but close, because she'd tried to use an AI chatbot to help her write a complaint letter to her insurance company, and it had hallucinated a policy number that didn't exist, and she'd sent the letter before she noticed.

She felt stupid. That's the part that got me. She kept saying "I should have known better" and "I'm not good with this stuff" and I was sitting there thinking, no, that's not right. You're not bad at AI. Nobody taught you how AI actually works. There's a difference.

So I went over to her place that Saturday with my laptop and we spent like three hours going through it together. What these tools actually are. What they're good at. What they'll confidently make up if you let them. How to check their work. And by the end of it she was using it to help her prep for a job interview she had the next week, and she was doing it well. She knew when to trust it and when to verify. She was having fun with it.

That changed something in me.

Because I'd been in AI-obsessed corners of the internet for years at that point. Following the researchers on Twitter, reading the papers I could halfway understand, testing every new tool the day it dropped. I had this whole internal framework for thinking about these tools that I'd built up slowly over time. And I'd just... assumed everyone else was building the same framework. I didn't realize how different my relationship with this technology was from people who hadn't been living in it.

After my mom, it was her coworker. Then my college roommate. Then my cousin. Then my cousin's husband. At some point I made a group chat called "AI stuff" and started posting tips in there, and people would share it with their friends, and then random people I'd never met were texting me asking if I could do what I'd done for my mom but for them.

That's when I thought, okay, maybe I should just make this a thing.

Here's what I've learned from teaching AI to people who don't come from tech: the problem is almost never intelligence. Every single person I've worked with has figured this stuff out once someone explained it in plain language without assuming they already knew things. The problem is jargon. The problem is tools that are weirdly intimidating even when they don't need to be. The problem is that most of the content about AI is written for people who already understand AI, and if you're not in that club, you feel like you stumbled into a conversation that started without you.

I'm not a researcher. I don't have a computer science degree. I studied communications, which is maybe the most ironic background for someone who now spends a lot of time trying to communicate technical things clearly. But I think that's actually why this works. I learned all of this as an outsider too. I remember what it felt like to not understand what a large language model was. I remember the first time I asked a chatbot something and it gave me a confident wrong answer and I believed it. I made those mistakes. I can help other people skip them.

The other thing that drove me to start doing this more intentionally: I was getting frustrated watching people get burned by AI in ways that were so preventable. Not because the tools are bad. I genuinely love a lot of these tools, but because people were going in with the wrong mental model. They thought AI was like a search engine that gives you complete sentences instead of links. So they treated it like an authority instead of a collaborator. They didn't verify. They didn't push back. They didn't know they could ask it to explain its reasoning or tell them what it was uncertain about.

And then when it got something wrong, which it will, that's just reality, they'd swing to the other extreme. "AI is useless." "I can't trust any of this." Which is also wrong! These tools are genuinely useful for so many things. You just have to know how they work well enough to use them well.

I want people to land somewhere in the middle. Appropriately skeptical. Genuinely curious. Confident enough to experiment and figure out what works for their actual life and their actual job and their actual problems.

I'm not trying to make everyone into an AI power user. My mom doesn't need to know how transformers work. She needs to know what to do when the AI says something that doesn't seem right. She needs to know how to write a prompt that gets her what she actually needs instead of a generic non-answer. She needs to know there's a difference between a tool being bad and a tool being misused.

That's the gap I'm trying to fill. Not the "here's every new AI tool dropped this week" content, there's plenty of that, and some of it is great. More like: here's what I actually use. Here's what I actually think. Here's what burned me, and here's what surprised me in the good way. Here's what I'd tell my mom, or my roommate, or my cousin's husband who has been using the same AI tool for six months and still doesn't know he can ask it follow-up questions.

I don't always get it right. Sometimes I explain something in a way that's still confusing and I have to try again. Sometimes I'm enthusiastic about a tool and then three weeks later I'm like, okay, I oversold that. I try to be honest about when that happens.

But I think there's something valuable in someone who's genuinely excited about this stuff, who also genuinely cares about regular people getting something out of it, writing about it in a way that doesn't require you to already be in the know.

That's what I'm trying to do here. Whether it's working, I'll let you be the judge.