Key Takeaways

  • AI assistants are a top-three priority for both Apple and Google over the next 12 months.
  • Startup lead times have shrunk from 6 to 12 months down to 2 to 4 weeks as competitors clone working features with AI coding tools.
  • Building happens at machine speed, but customer learning and organizational adoption remain locked to human speed.
  • Apple faces a structural handicap in AI agents because effective agents require cross-silo cloud data, while Apple is built around on-device processing.

The 14-Day Feature Half-Life

In previous software eras, a startup with a novel workflow had six to twelve months of breathing room. You found customer traction, iterated in quiet obscurity, and built a moat before incumbents reacted.

That buffer is gone. Today, coding assistants allow competitors to detect and clone a working product within days. As Jean-Denis Grèze points out: “The problem today is it is so much faster to build that as soon as something is working for somebody, everyone notices and is able to get there within like two weeks or four weeks like copy really really fast and learn.”

Speed of execution in pure code is no longer a defense. When everyone has access to the same frontier models and automated coding tools, building the software is cheap. The bottleneck moves entirely to distribution and understanding customer workflows.

Machine Speed vs Human Learning

The trap for modern founders is confusing rapid deployment with business progress. Grèze captures the core tension in one line: “You can build now at the speed of machines, but you can only learn at the speed of humans.”

You can generate ten thousand lines of code over a weekend. You cannot make an enterprise customer change their weekly workflow over a weekend. Customers still need time to onboard, test, get confused, give feedback, and build trust. Because software generation is instant, founders often outpace their own feedback loops. They ship five features before validating whether the first feature solved a real problem.

Worse, if your entire product relies on an edge case that frontier labs are actively targeting, your head start vanishes overnight. Grèze warns founders not to hide behind tactical advantages: “forget your distribution strategy, forget all like the fact that you're good at work, forget like the network effect features, but if codeex can do something that you cannot do and that thing is something that matters to users, it's over, right?”

Apple's Structural Cloud Deficit

Startups are not the only ones facing pressure. Big Tech is pouring everything into the space. Grèze notes: “I know what I'm building is a top three priority at Google and Apple like in the next 12 months. Not a top 10 priority, like a top three priority.”

Yet even with unlimited capital, Big Tech faces architectural constraints. Apple, for example, is constrained by its history of on-device computation and consumer privacy. Grèze observes that real enterprise agents require access to fragmented data sources across multiple cloud platforms: “And the reason why that matters is because what we talked about earlier, agents are better the more data they have. And the data is not all on the phone. And and so the fact that they're not cloud is like one big issue.”

If an agent cannot tap into distributed enterprise cloud data, it cannot automate complex work, no matter how good the on-device chip is.

What to Do With This

Review your engineering sprint backlog tomorrow morning. Flag any feature whose primary defense is that it took your team two months to code. If a competitor using frontier tools can clone it in two weeks, cut it or redesign it around proprietary enterprise data integrations that cannot be copied from a user interface alone.