Wide eyes, spinning minds, and bright sunshine

Reflecting on my train home, that combination feels like it captures the essence of our second AI x Design event this morning.

A huge thanks to Rich from Lovable and Benjamin from Monzo Bank for anchoring the session with two very different perspectives on where design practice is heading.

Here’s how I’d summarise the themes we covered:

1️⃣ Designing for agents, and agentic workflows

Rich’s talk leant heavily into how Lovable are scaling design quality through agent workflows (and a healthy dose of clearly written markdown files). But this theme came through in the conversations and demos beforehand too. People seem to be experimenting with agents far more than even a month ago. Workflows that gather and package up context before ‘design’ begins, and then evaluate and iterate on what’s built afterwards. A shift from our previous morning where the focus was more squarely on prototyping and shipping design changes to production.

2️⃣ When AI can generate everything, lean into knowing where to cut

Benjamin’s talk focused on building a visual world through art direction, product clarity and strong UX principles. And then bringing discipline and rigour to cutting away everything that doesn’t speak to the essence of the experience. In other words ‘taste’ shows up in what you remove just as much as it does in what you add.

Taste is such a buzzword right now. But Benjamin’s take made this feel less abstract, and more about honing your own ability to cut through the noise, quite literally, by reducing parts of an existing experience. Even when an LLM’s response is perfectly polished.

3️⃣ The demos were a brilliant patchwork of possibilities

Following the 90-minute build session we saw a bunch of cool projects. Too many to call out, but a few that stuck in my head: → A spatial desk allocator that guides you to your hot desk seat when you arrive to the office that day. → A geo-shader ‘talking orb’ as a UI for weekly reflections and coaching. → A product manager ‘stand-off’ tool where three PMs create independent PRDs and resolve trade-offs collaboratively, leaning into their individual personality quirks and taste. → Oh and a hyper-local ‘what to wear’ app for people new to London, using feels-like temperature as the baseline, and individual personality traits like people who naturally run hot or cold. Not just the forecast.

Simon and I are already plotting the next one. If you, or a designer you love, is doing wildly interesting things with AI, drop their name or project in the comments. We’re curious to find our next featured guests.