The number of mobile apps available on Apple’s App Store was up 30 percent last year, according to one estimate. But downloads were up only 3 percent.
It isn’t just apps. Software in general — everything made of code — is suddenly available in huge, heaping quantities. Thanks to artificial intelligence, you can now, in plain language, tell a computer to write code, in plain language, and it will often do an excellent job. These capabilities get better nearly every week.
But when was the last time you ended up marveling at a new app or game on your phone? Whither the applications like Slack that take industries by storm, or games like Candy Crush that dominate our subway commutes? Is it just all app slop from here on? Did A.I. kill the “killer app”?
That would be a natural conclusion to draw, but I think that we’re headed for a very different kind of digital revolution. We shouldn’t expect a spate of A.I.-made software to replace all our existing tools. After all, people don’t like to switch to new software unless they must. A.I.’s real strength may be to make many new things that play a more moderate, specific role: reducing sources of small friction that otherwise stack up in our lives.
Most big corporate efforts to use A.I. to increase productivity fail — perhaps as many as 95 percent, according to an M.I.T. study published last year. A boss sits a team down — let’s say it’s a financial services firm, or a consulting firm — and asks them to use A.I. to improve the sales process or track relationships, and after a lot of trial and error, it doesn’t work out. This finding is quite awkward, given the projected multibillion-dollar valuations of A.I. companies, but nothing I’ve learned from talking to dozens of people in the industry disproves it. People are hopeful that a bold A.I.-powered future is right around the corner. They are tantalized. They often use A.I. tools all the time individually. But companies’ internal efforts don’t work.
I don’t blame the coders for this, or even the bosses. I blame the big A.I. companies, who are in such a hurry that they often release new versions of their models and tools without manuals or even clear YouTube tutorials. When projects fail, the implication is that humans just need to stop everything else they’re doing and prompt the models harder.
But that’s a hard trick for an old industry to learn. The tech industry loves to act as if it just sprang out of Zeus’ forehead, but it’s actually a septuagenarian, and surprisingly set in its ways. It will take years before A.I. tools can fully supersede traditional approaches to making software. As dramatic as everyone would like this revolution to be, it’s barely begun.
Venture capital money that drives tech innovation is being poured almost exclusively into A.I. Products that aren’t A.I.-first (such as an old-school game or a task organizer) seem like a risky investment. Why invest in a culture that will soon be gone? If A.I. makes software cheap and abundant, and digital products are easy to clone, then it will be hard to compete. Software used to be the ultimate “moat” — but now it’s starting to look like a commodity. That’s another reason the App Store is now flooded with unimaginative apps instead of thrilling new efforts.
Which gets to something fundamental about A.I. coding, and A.I. in general: It’s very good at reproducing things that look and feel like what came before. It can make a PowerPoint presentation, or even, with time and coaching from a human, a flimsy version of PowerPoint itself. But making something new, something no one has seen before, takes much more time — sometimes just as long as it took before A.I.
Where A.I. excels, I think, is in letting you get access to aspects of computing that were previously out of reach. You can use A.I. to develop complex business models, machine learning systems or advanced software that runs in a web browser. These were all utterly normal things before large language models showed up, but they required specific skills and tended to be the domain of very highly paid software developers. A.I. lets you learn these things as you go, and you’re less likely to give up along the way.
The way I know this is true is that I see a lot of new software that only modestly improves certain things. Librarians are using A.I. to build better, more accessible historical archives; novelists are “writing” subsets of code and contributing to open-source projects; underfunded health care start-ups are whipping together custom tools to help people deal with insurance companies. None of this stuff ends up on the homepage of the biggest tech news outlets, and most of it probably will not make anyone a dime. It’s extremely normal software, not revolutionary or transformative. I think we’re witnessing a cultural throat-clearing — people scribbling in their notebooks, getting ideas out.
For a while, I must admit, it looked as if software developer roles like mine were done for. How could we fight against tireless robots? But our industry is slowly realizing 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. A.I. can write very good software, but 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.
The reason they shouldn’t is not spiritual but financial: Craft makes money, and A.I. tends to wreck your sense of craft. For months, we’ve seen report after report of companies encouraging all technical employees to use A.I. as much as possible to write more code — “tokenmaxxing”— even if costs skyrocketed. Now managers are hitting the brakes hard. The call to stop burning money? That’s a sense of discipline re-emerging.
Some of the most interesting conversations I’ve had about A.I. are with organizations like food banks and small museums who’ve wanted to build simple custom apps for their patrons and donors, clean up their data or create smarter marketing tools, but never had the money. Often they have real qualms about A.I. and its social costs, and sometimes they decide not to use the technology at all. But many want to deliver good things to their constituents, and that seems more possible now because of A.I.
The future may not be a few huge apps running on a few huge platforms, generating tons of revenue; it could instead be tons of little ones, purpose-built for the church, mutual aid group, company department or softball league. I’d enjoy a future in which the answer to “where’s the great software A.I. was supposed to bring us?” is: “everywhere.”
Paul Ford is a contributing writer in Opinion. He is a founder and the president of Aboard, an A.I.-powered software acceleration platform, and a co-host of “The Aboard Podcast.”
The Times is committed to publishing a diversity of letters to the editor. We’d like to hear what you think about this or any of our articles. Here are some tips. And here’s our email: [email protected].
Follow the New York Times Opinion section on Facebook, Instagram, TikTok, Bluesky, WhatsApp and Threads.
The post A.I. Was Supposed to Give Us New Killer Apps. What Happened? appeared first on New York Times.
