I was struck using Claude recently to ideate on a concept that it was so keen to build me something I had to resist.
Spending a lifetime building software I've found that building has gone from being one of the hardest parts to one of the easiest.
I'm immensely excited about the state of software development — for small teams and individuals, we can do more than ever.
But while it has become so much easier to build, it doesn't mean that everything that can be built should be built.
In a healthy team, when building was harder, more time consuming, and more expensive, it forced a level of thought and rigour on the decision and planning of every feature and certainly every product.
Now, though, building is trivial, fast, and almost free. So almost anything can be built almost instantly.
That's not always a good thing!
My recent discussion with Claude last week involved ideating on various methods to help me with managing our pipeline of prospects and customers for EcoSend. The responses from Claude were very helpful — with useful insights on industry-standard approaches for following up, pipeline stages, and overall pipeline management.
But Claude was all-too-keen to jump straight into building me a whole system for managing everything. A new dashboard, a new data store, a new structure, a new set of views.
Most of my time working with other people in a team has taught me that simplicity is extraordinarily difficult to maintain. And that you're continuously fighting tool fatigue. More more more is always the answer. Until you're drowning in tools and systems that no one can fully make sense of.
Every tool requires learning. Every tool needs to connect to another. Every tool must be maintained. Every tool might expose a new security threat. And in a team, you just need to multiply all of these issues by the number of people you work with.
For many individuals, the new power to build dashboards, websites, and apps is incredibly exciting and empowering.
But for anyone building a team, and in many cases, for the individual too, there is a hidden cost to every new system and tool. And that cost can be far more than the monthly price tag on the plans page.
So next time Claude is pushing you to build a new tool or system, it might be worth asking first: do I need a new tool? Or do I just need to sit a think?
I'd be willing to bet the latter might not hurt, regardless of the ultimate answer.