---
title: Where design is heading
slug: where-design-is-heading
summary: Design is opening up in both directions, up into business strategy and down into agents, model behaviour and API design. We're in the messy middle, and knowing which problem is worth solving is still what separates good from great.
url: https://charlieholland.com/
---

# Where design is heading

A recorded design interview, published 24 August 2026 (about 15 minutes to watch). It plays in the Perspectives section of charlieholland.com. Tags: Product Design, Artificial Intelligence, Careers, Opinion.

This is the fullest statement of my views on where design is going. The through line is that the layers of abstraction between design, product and engineering are condensing, and that the future belongs to what I call the polymath, someone with a broad shape who can still go deep in a few areas. That does not mean every designer should be architecting a backend. It means being able to lean in, build something in code, test it quickly and move.

The practical version is that the vertical has opened up in both directions. Designers can move up into business strategy and automation, or down into the technical layer, designing and evaluating agent behaviour, understanding how large language models actually work, and doing the trust and transparency work that sits on top of them. Evaluating agents is close to a new form of usability testing. What made the technical direction hard before was access, not capability. Understanding the shape of a backend used to mean an hour-long meeting with the data engineers. With tools like Cursor and Claude Code you can query it directly, understand the API shapes and responses, and collaborate far faster. That is why I think API design is becoming a genuine design surface, and OpenAI is already hiring for exactly that hybrid profile.

On whether design still matters, I think the market has answered. OpenAI acquired Jony Ive's company for around six billion dollars, and Anthropic is hiring designers from a wide range of backgrounds. My thesis is that it is going to be an extraordinary time to be a designer. We are in what I call the messy middle, working out how design is best used in this new age, and that resolves over the next couple of years rather than the next six months.

What organisations need to do is create the space. The pattern I like is the small dedicated team, the AI centre of excellence or innovation hub that can move quickly, test and validate ideas and find where AI genuinely earns its place. That is the same shape as DARPA or Skunk Works, and it connects to the argument in my article on the decline in innovation and the lost power of observation.

I think design brings three things to that. Designers are good at future vision, at seeing where things are heading and working back from it. They are obsessive about the customer and can translate that into what matters to the business. And they are the glue between the business side and the technical side. Put those together with real technical understanding and it is a compelling profile. Some designers would rather stay in their lane and go very deep, and that is a legitimate choice, but I am opinionated about this one.

The skill that matters most has not changed, and it is the softer one. Problem framing. Being clear about which problems are worth solving now versus later. Now that you can design and build things so quickly, it is easy to lose the forest for the trees. What separates good from great is what I call the golden thread, running from the strategy and the business context, into the problems actually worth solving, and out to a real return on investment. I see people fall short in interviews for exactly this reason, because they focus too much on the solution.

Watch it in the Perspectives section of https://charlieholland.com/