The baseline has changed

Think back to when you got your first computer and started writing code. That sense of wonder when “Hello World” appeared on your screen. The dizzying feeling that you could make this strange machine do what you told it to. This feeling is useful again. It’s much better to mess around with these new AI tools and get your hands dirty building stuff than to sit there having opinions about them.

If you approach these tools like that, a lot of the online AI developer discourse starts to look beside the point. The mistake people make with AI tools is treating them as a debate topic. Hype, backlash, benchmarks, model tribalism. None of that is the main thing. The main thing is that the baseline changed.

What felt like impressive software output two years ago can now be done much faster. Tasks that used to take days can often be pushed through in hours. Prototypes appear faster. Refactors get cheaper. Throwaway software to solve annoying problems is now viable. The cost of getting to a plausible implementation has collapsed.

That does not mean every team is suddenly great or all software delivery got easy. It does mean expectations are changing, and what feels “fast enough” is no longer the same. Many organisations are underestimating what’s happening. They are still comparing today’s tools to the old normal. The game is figuring out ways to use the new tools to move faster without sacrificing quality. It’s a new way of working, solving problems and eliminating bottlenecks in the entire stack, from product and developer tooling to infrastructure.

That is the shift. Not just new tooling.

New expectations.