Easier Than Ever to Build the Wrong Thing
AI lets teams skip the "should we build this?" conversation. On my own projects, building is how I have it.
In the age of AI it’s incredibly easy to build at pace. Does that remove the discipline people need to apply? They’re skipping the thinking, the discussion and the ideation around “should we build this thing?” because it’s so easy to just start building it and shipping it.
In large software teams this is surely causing a tonne of feature bloat. It’s not necessarily going to be Frankenstein software, but it’s not focused on the end user or what people are actually trying to achieve. And once something’s in production, as we know, it becomes really difficult to take away.
The teams that end up best in this day and age will have strong product leadership that can harness this, not just build unscrupulously and quickly. Keep the thinking strong. Get in the room, have the brainstorm, throw the ideas out (the human ideas, not just the AI ideas), and get your hypotheses out:
- what problem you’re trying to solve
- which user group you’re trying to fix it for
- what people are trying to achieve
- what superpower you’re trying to give the user
That skill hasn’t gone away, and AI isn’t good at determining it with the context it has.
Then again, on my own stuff, building is the thinking. I’m not proclaiming to have hit some higher power level of productivity. I’m simply in awe that I can tinker with ideas where I’d previously have sat wondering whether it was worth convincing a developer friend to build it, and recruiting a CTO, before I could make any progress. Now I can build the literal actual thing and realise that… no, it’s not worth investing in. But what a time to be alive.