Is the end of Excel near?

Not Excel per se, but Excel as the universal tool for getting an overview of some kind of data.

This specifically struck me when I saw our (high) water bill for 2025 and thought about tracking the meter reading on a weekly basis. You always do that with Excel. It’s not pretty, you have to read the numbers, enter them into Excel, but then you can see the trend, calculate the average, etc.

This kind of Excel usage is (at least in my world) omnipresent. Whenever you want to collect some data, you fire up Excel. Even if it’s no fun. But it works.

This time, instead, I briefly asked the LLM (specifically: Opus 4.5 in Cursor) to write an HTML+JS+node.js backend app, deployed the app on my server, and just like that I had a much nicer app where it’s almost fun to enter and monitor meter readings. The whole thing cost me 15 minutes to write the spec and 5 minutes to copy it over and integrate it into pm2.

Here are three partial screenshots of the app on my phone. It really is much prettier than Excel.

Now, the part about running your own server and deploying apps on it is not exactly pleasant. And you’d probably prefer a native app. But that’s exactly what the LLM providers are working on right now, for example OpenAI’s development of an app directory.

There are also providers – such as the brilliant render.com, or the AI-based bolt.new from Stackblitz.com – that make it super easy to deploy your own apps to a server. (And the good old Replit, whose deployments always gave me trouble back in the day.)

I’m convinced that by the end of this year, personal apps that you can have an AI write for you, completely tailored to your personal needs instead of “off the rack”, will be normal. And many such personal apps will be nicer, more practical, simpler versions of what we currently do with Excel because there’s nothing better – even though Excel is probably not much fun for most people.