0
0
For a couple of years now, the promise has been simple: you don’t need to be a programmer anymore. Describe the app you want, and an AI will build it for you.
So people did. Lots of people. And now one of the best-known writers on software is asking an awkward question out loud: if anyone can make an app, where are all the AI killer apps?
The answer turns out to be less about technology than about something much older, and much more human.
Where Are the AI Killer Apps? The Numbers Behind the Question
Paul Ford has spent decades building software and writing about it. On September 12, 2026, he put the question plainly on social media: “If these things can code (and they can!) then why aren’t we seeing any great software?”
His column that same day, in a major US newspaper, pointed to a gap that is hard to ignore. As the newsletter Numlock summarised his figures, the number of apps available on Apple’s App Store grew by about 30 percent last year. Downloads grew by about 3 percent.
Read that again. Growth in apps outran growth in downloads roughly ten to one.
And the flood hasn’t slowed. The app-tracking firm Appfigures found that new app releases in the first quarter of 2026 were up 60 percent on the year before, across both Apple’s and Google’s stores, as reported by TechCrunch. On Apple’s store alone, the jump was 80 percent. Those figures count releases, not success. Nobody is claiming the new apps are being used. That’s exactly the point.
Making things got easy. Wanting things did not.
This Isn’t a Sceptic Saying “I Told You So”
Here’s the detail that makes Ford’s question worth taking seriously. He isn’t an AI doubter.
Back in February 2026, writing in the same newspaper, he admitted he was excited about AI coding tools, even though most people around him weren’t. “All of the people I love hate this stuff, and all the people I hate love it,” he wrote. “And yet, likely because of the same personality flaws that drew me to technology in the first place, I am annoyingly excited.”
So this is an enthusiast checking back in seven months later. And he still says the tools are good. His words: “A.I. can write very good software.”
But then comes the part that stings: “it also makes it easy to do someone else’s job badly, which is part of why all those projects fail. Now that everyone can code, it’s become clearer why many shouldn’t.”
That sounds harsh. It’s actually hopeful, especially if you work in any skilled job. Ford writes that “for a while, I must admit, it looked as if software developer roles like mine were done for.” What changed his mind was noticing that “making truly cutting-edge software still requires humans to think and work together, to maximize their skill sets and to practice their respective crafts.”
In other words, the machine can type. It can’t care. And it can’t sit in a room with other people working out what is actually worth making.
Why More Apps Don’t Mean Better Apps
There’s a simple reason the pile of apps grows faster than the number of people using them. Nobody gets more hours in a day because apps got cheaper to build.
Everyone on earth has the same 24 hours. The same tired eyes at night. When making something costs almost nothing, the thing in short supply is no longer the product. It’s a person’s attention, and a real reason for them to care.
That changes the question every maker has to ask. It used to be “Can I build this?” Now the answer to that is nearly always yes. The harder question is “Who is this for, and why would it matter to them?”
You can see the same squeeze in the workplace. When AI speeds up a task, the saved time rarely comes back to the worker. Often, AI ends up adding to the workload instead of lightening it. And a large survey of executives found that the promised productivity gains mostly haven’t shown up yet. More output, it turns out, isn’t the same as more value.
What This Means If You Make Anything at All
You don’t have to write code for this to apply to you. It fits anyone who cooks, teaches, writes, sells, fixes, or builds. A few questions are worth carrying:
- Start with a person, not a product. Picture one real human you know who has the problem. If you can’t picture them, the thing probably isn’t needed yet.
- Ask what you’re adding to the pile. The world has plenty of “fine.” What does yours do that nothing else does?
- Don’t do someone else’s job badly. If a tool lets you skip a skill, find someone who has that skill and work with them. Ford’s point was that great work still comes from people working together.
- Count use, not output. One thing someone keeps using is worth more than ten things nobody opens.
None of this costs money. It just costs a little honesty before you start.
An Old Complaint About Too Much Making
Here’s what surprised me. This complaint is ancient.
An ancient book of wisdom ends with a tired writer saying, in effect, that there is no end to the making of books, and that all that study wears a body out. That was written long before the printing press, never mind apps. Even then, making had already outrun anyone’s ability to take it all in.
What’s striking is where that writer goes next. Not to a faster way to make more. Not to a better filter for the pile. The book steps back and asks what the whole thing was for, and finds the answer not in the output at all but in God. It’s a quiet reminder that a person was never meant to be measured by how much they produce.
Which may be the real relief hiding inside Ford’s question. If machines can now make endless things, then making endless things was never what made us valuable.
The Next Killer App Will Probably Start Small
Nobody knows what the first great AI-built app will be. Ford’s column points toward people thinking and working together, practising a craft. It’s hard to picture it coming from a lone prompt typed at midnight.
If you’re building something right now, that’s good news. The part of the work that matters most is still yours: noticing a real person, caring about a real need, and choosing to make one thing well instead of ten things fast.
The pile will keep growing. What people will always need is something made by someone who actually cared.
Discussion Question
If anyone can now build an app in an afternoon, what do you think will make one actually worth using? Is it the idea, the craft, or the people behind it? Tell us what you think in the comments.
Share This
App Store apps grew about 30% last year. Downloads grew about 3%. Making things got easy. Getting anyone to care didn’t. https://bgodinspired.com/?p=115595
Paul Ford, who was “annoyingly excited” about AI coding in February, now says AI “makes it easy to do someone else’s job badly.” Worth reading if you build anything: https://bgodinspired.com/?p=115595
The thing in short supply isn’t apps anymore. It’s a reason to care. This one made me rethink how I measure my own work. https://bgodinspired.com/?p=115595
Questions People Are Asking
Why aren’t there any killer apps made with AI yet?
Software writer Paul Ford raised this question in September 2026, noting that AI tools can write very good code but that great software still requires people to think and work together and practise their crafts. In his view, AI also makes it easy for someone to do another person’s job badly, which helps explain why many AI-built projects fail. Building an app has become easy; building one that people genuinely need has not.
How much has the number of apps grown because of AI?
According to figures in Paul Ford’s September 2026 newspaper column, as summarised by the newsletter Numlock, the number of apps available on Apple’s App Store grew by about 30 percent in the previous year while downloads grew by only about 3 percent. Separately, the app-tracking firm Appfigures reported that new app releases in the first quarter of 2026 were up 60 percent year over year across Apple’s App Store and Google Play, and up 80 percent on Apple’s store alone.
Are more apps being downloaded now that AI makes them easier to build?
Not in proportion. Figures cited in Paul Ford’s September 2026 column showed App Store downloads growing about 3 percent in a year when the number of available apps grew about 30 percent. Appfigures’ 2026 data on the surge in app releases measures how many apps were published, not how many were used, so a rise in releases does not show a rise in demand.
Will AI replace software developers?
Paul Ford, a longtime software developer and writer, said in September 2026 that for a while it looked as if roles like his were finished, but that the industry is realising cutting-edge software still requires humans working together and practising their crafts. AI can write good code, but deciding what is worth building, and building it well for real people, still depends on skilled human judgment and collaboration.
