Thirty Years of Squinting at Menus

There's a small thing that drives me a little crazy, and I'd bet you've felt it too.

You're hungry, you pull up a restaurant on your phone, and you tap the menu. And you get a PDF. So now you're pinching and zooming, dragging a tiny page around with your thumb to read whether the salmon is sixteen or twenty-six dollars. It's a beautiful menu. Someone was paid good money to design it. And it's almost unusable on the one screen most people will ever see it on.

We've been doing this for thirty years. The web has changed more times than I can count, and somehow the menu on your phone is still a photograph of a piece of paper.

I want to be clear that this is nobody's fault. This is the spot where a lazier writer would call restaurant owners behind the times. I won't, because they're not.

Here's what actually happened. A good menu changes constantly: the chef finds a great piece of fish one week, a favorite ingredient goes out of season the next. Turning each new version into a real web page used to mean coordinating a designer and a web person on a timeline that never matched the kitchen, at a cost that never made sense for a dish that might be gone by Friday. So owners did the only sane thing. They had the menu designed once and posted the PDF. Stale or expensive, pick one. Most people picked stale, and honestly, who could blame them.

This is the part of the AI story I find worth talking about, and it has nothing to do with robots.

For thirty years, fixing this small thing cost more than the annoyance was worth. That math just flipped. Now a restaurant owner can text us the new PDF, the same way you'd text a photo to a friend, and a little while later their website has a real menu on it. Kai reads the PDF and turns it into a real web page that's easy to read on a phone, no pinching, no squinting. The owner didn't open a website builder or wait three days for a developer. They sent a text and went back to the kitchen.

Once you notice this kind of fix, you start seeing the same shape everywhere.

Think about the website that's quietly wrong. A florist changes her hours for a holiday, or closes early for her kid's recital, but the site still says open, so somebody drives across town to a dark shop. For years, fixing that meant logging into some clunky site you'd half forgotten the password to, so most people just didn't, and the site slowly drifted away from the truth. Same trap as the menu: the fix cost more attention than the problem seemed to deserve. Now she texts the change and the site is right within minutes.

None of this is dramatic. Nobody's going to write a breathless headline about a menu that's easy to read. But this is what the useful part of new technology has always looked like once the hype burns off.

Until one day you realize you haven't pinched and zoomed a menu in months, and you can't remember when that stopped.

That's the part I'm betting on. Not the technology that promises to change everything. The kind that quietly fixes the little things we all gave up on.