AI Hype Has Become Unbearable

If I see another thumbnail saying “This changes everything” an hour after a new AI model comes out, I’m going to lose it.

There’s a lot of interesting technology to learn. But telling people they’re being left behind unless they learn another 500 tools isn’t getting them excited. It’s making them want to leave the industry.

I’m saying this as someone who teaches this stuff and wants to keep learning it. I want to see what people are building and understand how these tools work. But I’m tired of digging through the hype just to find out what a tool actually does.

I talked about this in AI Hype Has Become Unbearable on YouTube, because the way we present this technology affects whether people want to learn it at all.

Every release can’t change everything

What prompted this was Jev, a newly released model that interested me. It’s built to return structured decisions that software can use directly, rather than generate a text response. You give it context and specific questions, and it returns things like a choice between options or a score, along with probabilities. That makes it interesting for tasks like routing a request to the right place in a workflow. I want to explore where that approach is useful in practice.

I wanted to spend some time with it and potentially make a crash course once I understood it. But the reaction on YouTube, X, and blogs was already turning me off.

When every release changes everything, eventually you stop caring. It becomes the boy who cried wolf. The technology might be worth learning, but you’re rolling your eyes before you even click play.

That also makes it harder for people trying to teach without the hype. I’ve made crash courses on Claude Code, Cursor, and OpenClaw, along with my Blueprint workflow video. Those were straightforward Traversy Media tutorials: showing you how to use something.

Sometimes those videos get lumped in with the exact content I’m frustrated by. People see AI in the title and immediately tune out. I understand the fatigue, but I’d ask people to judge the actual content. There’s a difference between teaching someone a workflow and declaring that their entire career changed this morning.

I’m not anti-AI. I’m tired of the hype around it.

Show me what happens when something needs to change

The launch-day videos often follow the same format. A model comes out, someone gives it a prompt, it builds something in one shot, and we skip to the finished result. Then comes a benchmark and a lot of excitement.

What I want to see is what happens when you need to change something.

Put the tool into an existing project with bugs and decisions made six months ago. Show how it handles the change, where it gets confused, and what you have to do to get the work finished. That’s where I’d actually be using it.

A one-shot demo can be interesting. It just doesn’t tell me enough to judge how useful the tool will be in my own work.

If a model came out an hour ago, you can share your first impressions. But saying it changes everything takes experience and actual use cases. That’s why I don’t rush to teach brand-new technology. I need to use it before I feel comfortable explaining it to someone else.

My OpenClaw crash course came out six months after its release because I needed that time with it. By then, the hype had died down, but I had something useful to teach.

I understand why creators are struggling

I’ve seen people whose work I’ve enjoyed for years get pulled into this style of content. I’ve probably made videos that could be perceived that way myself, so I’m not putting myself on a pedestal.

I also don’t suddenly lose respect for established educators because they’re trying to find their footing. People like Max from Academind, Mosh, and Net Ninja deserve respect for the work they’ve put into helping people learn.

We built learn-to-code platforms from nothing and put our hearts into them. Now the industry is changing, and it can feel like the thing we worked so hard to build is being phased out. I understand why people don’t know where to go next.

The first two years of my YouTube channel, I made absolutely nothing. I did it because I loved it. Even though it eventually became a business, teaching was the reason I started.

I’m still figuring out where I fit into all this, too. I want to keep teaching useful things without feeling like every video needs an exaggerated claim to get someone’s attention.

You don’t have to replace a workflow that works

Social media makes it look like you have to use every new model immediately and rebuild your workflow every week.

If what you’re using works, keep using it.

When something catches your attention, try it on a task you actually need to do. See whether it helps. If it does, add it to your workflow. You don’t need to adopt it just because everyone is talking about it.

Take benchmarks with a grain of salt, too. A model generating some weird game better than another model doesn’t necessarily make it more useful for your projects. What matters is how it performs on the work you need help with.

I don’t want the marketing to turn people off AI altogether. There’s plenty worth exploring. But you can be interested without treating every release as an emergency.

What I want to keep teaching

Using AI well involves more than opening a prompt and saying, “Build me an app.”

There’s context management, skills, subagents, code review, CI/CD, and how you put those things together into a workflow. There’s a lot of technical material to teach.

It’s harder to turn that work into a video than a conventional coding tutorial. When I’ve written the code beforehand, I know what’s going to happen. With AI, there’s waiting, and the results can be unpredictable. I think that’s part of why so much content falls back on quick demos.

But it is possible to teach this stuff properly. I’ve made a 16-hour coding-with-AI course, and the response from people taking it has been encouraging. I want to do more of that kind of work.

I also want to keep teaching coding. I love it, and I still believe people need the fundamentals. The format may change. Shorter, focused tutorials around particular concepts make sense to me, with longer courses for larger projects. You need to understand what you’re building, even if you don’t memorize every bit of syntax.

If I’m excited about a tool, I want to explain why and show where I actually use it. If it falls short, I want to show that, too. There’s enough interesting technology out there without exaggerating what it can do.