Stop waiting on each other
The roles of design, engineering, and product are converging - and most people are drawing the wrong conclusion from that.
The roles of design, engineering, and product are converging - and most people are drawing the wrong conclusion from that.
- Convergence means overlap, not replacement
Working closer to code has made me appreciate more, not less, of what engineers do. I'm not going to debug an incident or improve core infrastructure. Engineers hold a holistic view of their domain, sweating details that are invisible when you're moving fast. Same goes for designers and product people.
The wrong conclusion from convergence is "now I can do your job."
The right one is "now we can stop waiting on each other."
- Gather context at speed, prototype, then discuss.
Before pulling the team together, I synthesise what customers have been saying about a problem, and cross-reference it against competitors and what we've already built. I use Claude Cowork connected to Linear and our codebase to do this quickly.
Based on the context gathered, I build a working prototype. Not a static Figma one with a million nodes to manage manually. A flow that has realistic demo data. If it's a small task, I can also generate designs based on the codebase with Pencil.dev.
It's not perfect - we're still figuring out:
- Connecting your design system so you don't reinvent patterns
- Communicating the concept clearly to stakeholders. If it's a customer, getting insight on what's most important to solve is still key!
- Ship small wins instead of managing tickets
I'm on an ambitious team with many projects in progress at all times. These days, I use AI coding tools like Conductor and Cursor to answer questions about the codebase, ship a quick fix based on feedback, and unblock myself.
These things would have otherwise sat on an engineer's to-do list. Need product analytics events? I create them myself. Bad copy? Updated. These reps build the confidence to ship bigger improvements.
It's not about replacing engineers - designers shouldn’t feel pressured to contribute production code - it's about owning your impact.
If you're a designer who wants to start
I think the easiest way is to build something you care about with AI tools. I'm sure there are more methodical paths, but following what's fun worked better for me.
I started with v0, which helped me build the confidence to build my own web apps like this tarot prompt generator - and I'm just getting started.
You don't need to master a programming language first. Go as deep as you want - the point is to start.
If you get stuck, learning to troubleshoot or debug with a tool like Claude Code is literally what engineers do - you're building the same muscle.
On a different note: I've been building personal software that connects to my bank's API to analyse my own finances, and I'm unpacking why managing money is so unnecessarily painful.
So far, I think it's because we want to use money in alignment to our values and goals, but there's always a gap between the plan and what we do day to day.
What's annoying you about managing your money?
Reply or get in touch - I read every email.
Kina
