AI Gives Us a Chance to Rethink the Web

I’ve been thinking about how AI has already changed the way we interact with the web, and the interesting part isn’t the AI features getting added to the apps we already use. It’s that we have a chance to rethink the whole experience those apps sit inside.

It’s pretty amazing how much we’ve built since the early days of the web. But we’ve also wandered down some paths that made complete sense at the time and have quietly made being online exhausting. Ads and pages written for clicks (I’m looking at you Recipe blogs!), my inbox is mostly companies who would like me to buy something I already bought, and social feeds reward engagement, which in practice often means showing us whatever makes us angry enough to keep scrolling.

As someone who builds software, I understand exactly how we got here. Engagement matters, businesses genuinely need to know whether people are finding their products useful, and getting someone to come back is usually a good sign. The trouble is that we ended up designing the experience around the mechanics of measuring all that, and once the mechanics are the design, everything else has to fit around them.

AI gives us another option, because instead of making people navigate the web to get what they need, we can build experiences that understand what they’re actually trying to accomplish. Maybe I don’t need to open ten restaurant websites to find one with a table, or read every email to work out which three actually matter, or be given a feed that decides what I should care about today.

I’m not suggesting the internet as we know it should disappear. Websites, businesses, writers, and creators are still the reason any of this is worth using in the first place. What changes is the interface between all of that and the person on the other side of it, which is honestly the layer we’ve iterated on the least.

That’s the opportunity I keep coming back to. We can take everything we’ve learned from thirty years of building the web and revisit some of the choices we made along the way, not because the old ways were wrong, but because we finally have technology that can work differently.

The question I like to ask myself is what I would have built if I’d had AI back when I first started making websites, long before Google and long before social media. I keep thinking about all the afternoons I spent bug fixing Internet Explorer 6 issues, imagine if I never had to? Now we have a chance to rewind and build the internet we really wanted had we not had the constraints.