The Cost of Code Is Collapsing—But That Doesn’t Mean We Want More Apps
Over the last couple of years, something remarkable has happened: the cost of creating software has begun to collapse. AI-powered tools from companies like OpenAI, Google, Meta, and Wolfram have turned what used to be hours—or weeks—of programming into something closer to a conversation. You describe what you want, and the AI writes it for you.
This is a profound shift. It means software is no longer just the domain of trained developers. With the help of large language models (LLMs), anyone with a good idea and a rough sketch can generate working code, deploy apps, and automate tasks in minutes. We are, in a sense, entering an era where the cost of coding trends toward zero.
Naturally, this leads some people to talk about a future filled with “infinite apps”—a digital Cambrian explosion where new tools and utilities appear constantly, built on demand, tailored to niche use cases or even individual users.
But here’s the catch: humans don’t have infinite needs, infinite time, or infinite attention.
The Attention Bottleneck
Let’s start with the obvious. Most people today access the internet through their phones. And most of that time is spent on a small handful of apps: YouTube, TikTok, Instagram, messaging platforms. Our screens are already saturated, and our attention is maxed out.
Even if we could create a million new apps overnight (and technically, we almost can), users simply aren’t looking for more places to spend their time. They’re not craving ten more budgeting tools, twenty more to-do lists, or a hundred more weather apps. If anything, they want fewer apps—ones that do more, integrate better, and disappear into the background.
Businesses Have Limits, Too
You might argue that businesses could still benefit from all these new AI-generated tools—and many will. But even in the enterprise world, there are constraints:
- There are only so many employees to train.
- There are only so many processes that need digitizing.
- There are only so many hours in the workday.
Yes, AI will help companies create custom internal tools faster and more cheaply. But each new tool adds complexity—more interfaces to manage, more systems to integrate, more places where things can break. At some point, organizations don’t need more apps—they need better automation and smarter workflows.
Infinite Code, Finite Use
What’s actually happening isn’t so much a world of infinite apps, but a world where software becomes invisible. Instead of separate apps, AI agents will be woven into existing platforms, responding to natural language commands, quietly optimizing routines, and automating tasks without ever needing a new download or a new icon on your home screen.
We’ll see:
- Custom tools generated on demand for a specific task and discarded afterward.
- Bots that handle tasks across systems without needing user interfaces at all.
- Personal AI assistants that act as “meta-apps,” handling dozens of use cases behind a single conversational interface.
This isn’t about app abundance—it’s about app absorption.
The Future Might Be Different—for AIs
Now, here’s the speculative twist. Some people imagine a future where AI “individuals” or agents emerge—autonomous entities that act on their own behalf or on behalf of others, across networks, systems, and tasks. If this happens, we might truly see demand for massive amounts of software—apps not for humans, but for AIs.
If there are trillions of AI agents, each with unique goals, preferences, or decision-making models, then maybe we really would need something like “infinite” apps. Not for us—but for them.
But that world is not here yet. It remains more in the realm of science fiction than fact. For now, the bottlenecks are human: attention, usability, integration, and real-world value.
What We Should Really Expect
So what does the future look like?
- Yes, we’ll see an explosion of software creation.
- Yes, more people than ever will be empowered to automate, experiment, and build.
- But no, we’re not headed for a world of infinite apps in the traditional sense.
We’re headed for a world where the friction of software melts away—where apps aren’t things we download or open, but things that happen when we need them, triggered by a request, a pattern, or a prompt.
The real story isn’t about making more apps. It’s about using AI to reduce the need for apps in the first place.
Related Reading:
All comments are held for moderation. We'll only publish comments that are on topic and adhere to our Commenting Policy.