We're Engaged, But We're Not Participating

I switched from ChatGPT to Claude a few weeks ago, mostly because it felt like everyone was doing it, but also because I had a project I actually wanted to build. Not a conversation, not a back and forth of prompts and answers, but something with a website behind it. That distinction turned out to matter more than I expected.

The signing up part was easy. It always is. Pay, access, watch a tutorial, copy someone else's clever prompt from a thread you found at midnight. Consumption has been optimised to the point where you barely notice the steps involved. But the moment I actually tried to build something, I hit a wall that had nothing to do with the tool being bad and everything to do with what it was quietly asking of me. Tokens. Settings. Peak hours versus off-peak hours. How to structure a prompt so the model does what you mean instead of what you literally said. None of this is complicated once you know it. But nobody tells you any of it upfront. The ease that the AI revolution keeps promising doesn't survive contact with the canvas.

I don't think this is specific to AI tools, though they make the pattern obvious in a way most things don't. Access has never been easier. Sign-up flows have been refined for two decades. Checkout is one click. Onboarding screens exist specifically to remove friction before you've paid for anything or committed to anything. Entire product teams are employed to shave seconds off the time it takes someone to start consuming. Almost nobody has done the equivalent work for the moment someone tries to make something. The friction didn't disappear. It moved, from the front of the experience to the part most people never reach.

That relocation changes behaviour in a way that's easy to miss if engagement is the only thing being measured. Someone can sign up, log in, explore a tool for twenty minutes, and still create nothing, not because they lost interest but because the interface never actually told them how to begin. There's a real difference between engagement and participation, and most digital products are built to maximise the first without ever asking about the second. You can spend genuine time inside a platform and still be a passive presence the entire time you're there.

There's also a specific kind of loss in how these tools are built, beyond just the friction. Almost none of them are designed for you to try something, get it wrong, and learn from that in the moment. There's an expectation to arrive already knowing what you want and how to phrase it, execution from the first attempt rather than the kind of experimentation and play that usually precedes anything good. Trial and error is how people actually learn to make things. Most of these systems are built as if that stage doesn't exist, or shouldn't.

Cultural institutions run into a version of this the moment they try to build something beyond a browsing experience. A museum website that lets you scroll through a digital collection is a consumption product, and a well built one can be genuinely good at what it does. But the second an institution tries to build something that asks visitors to respond, contribute, or make something of their own, the difficulty jumps. The idea is usually fine. What's missing is the design work on what a creation on-ramp actually looks like, the equivalent of a one-click checkout for the act of making something rather than buying something. Most teams are building that from scratch, and if it isn't right, people bounce off it the way I nearly bounced off Claude, except they won't come back later to try a different prompt structure. They'll just close the tab and go back to scrolling the collection.

It would be convenient to explain all of this by saying people don't want to make things. Easier to stay in consumption mode, easier to watch, easier to keep scrolling past. But that's a description of the outcome, not an explanation of what caused it. We're not lazy. Creating is harder, and it's harder by design, whether or not anyone building these tools meant it that way. Every system that made consuming effortless left the creating side untouched, and untouched doesn't mean neutral. It means whoever built the thing spent their attention on the front door and none on what happens once someone is actually inside trying to do something with it.

I think about this every time I open a tool that promises creation will feel as easy as consumption, and then buries the actual creative act under a list of settings nobody explained first. The promise and the product don't match, and the gap between them is where most people quietly give up and go back to consuming instead. We're engaged, but we're not participating.