Tom and Jonny sit down with Sara Vienna, Chief Design Officer at Metalab, following her Hatch keynote on the trust gap. They dig into what designers can control in the AI race, unpack the tools Metalab's designers now build for every stage of a product, and get Sara's honest take on hiring and why nobody reads case studies.
Key takeaways
Designers can fight for trust, one sprint at a time. Sara's Hatch keynote, The Trust Gap, is about finding the gumption to take that fight to the C-suite. Even when your influence moves product by product, being pointed at a good fight is what lets you sleep at night.
Distrust of big institutions is old, and that's useful. People have doubted companies, governments and capitalism for years. Sara opened her talk with TIME's 'Trust Us' cover of Sam Altman and Greg Brockman and the teaser for Nathan Fielder's Elizabeth Holmes documentary, then moved on to what designers can actually do about it.
History is the best antidote to AI doom and hype. Sara looks to the Industrial Revolution, when people had to trust a metal contraption over a horse, and Henry Ford's cars came with real consequences, road deaths among them. Handmade things became more special once everything else was automated, which gives her hope for craft.
Use AI to make it real in week one, then let people make it remarkable. What used to be a 100-page deck and too many meetings can now become a tangible prototype that doubles as the PRD. That output is remixed from what's been done before, so Metalab runs it through a quality filter and a differentiation filter until it's worth shipping.
Designers can now build tools for the rest of the company. Sara's team makes tools for brand, creative and design systems, with quality control at every stage gate of a product's life. Her line about designers building tools for people outside the design team has become a staple of Tom and Jonny's tool-making workshops.
When anyone can make anything, a studio's value is its track record. Clients now arrive with something they've made themselves. Metalab treats it as a starting point, explains why it doesn't work and makes it great. Founders who'd rather ship it themselves are welcome back when they're ready.
Agencies are a great place to hone your craft right now. Senior designers stay at Metalab for the variety of problems. Some of its principal designers started as interns, and Sara is working with CEO Luke Des Cotes to rebuild that flywheel at scale, because juniors in a remote team need far more deliberate support.
Show what you've made, because nobody reads case studies. Sara hires for fire in the belly: people making things she's jealous of, with or without big logos in the portfolio. With AI, anyone can build something that solves a problem, or something just for fun.
Write a positioning statement for yourself, as if you were the client. Ask people you've worked with for their favourite and least favourite thing about working with you, unedited. Find the overlap between what people will pay you for, what you're good at and what brings you joy, then let your portfolio say it.
Interfaces for AI are still in their childhood. When the result set is effectively infinite, our boxes and lines can't show the richness of what comes back. Sara is more interested in interfaces that generate themselves to fit, which Metalab is starting to explore for clients.
Bring your most engaged users closer, and save the polish for the masses. Early adopters will tolerate some mess, so work with them in smaller, playful spaces. Metalab built Atlantic Labs to let readers explore The Atlantic's journalism with AI without touching the core product. Meanwhile, a design leader backed by the CEO and board keeps the main product from being enshittified.
Ask whether each feature still deserves to exist. Something that earned its place six months ago might not today. Agreeing whether you're more Beatles or Rolling Stones, Netflix or Apple, ends the debates that go round in circles.
Chapters
0:08 — Welcome, and meeting Sara at Hatch in Berlin
1:04 — Sara's Hatch keynote on trust, and a series of talks on our relationship with AI
4:37 — The 'Trust Us' TIME cover, the labs calling for a slowdown, and Elizabeth Holmes
6:13 — Jonny on holding excitement and fear about AI at the same time
8:47 — Optimism, enshittification and taking your fight to the C-suite
11:31 — What history teaches: the Industrial Revolution and Henry Ford's faster horses
15:00 — Craft after automation, and a hard question about Suno and Midjourney
15:50 — How Metalab decides which AI clients to take on
18:00 — How AI has changed Metalab's ways of working
21:11 — The tools Sara's team builds for brand, creative and design systems
22:04 — Designers making tools for people outside the design team
23:52 — Staying the expert when clients can make it themselves
25:41 — Why a premium service still makes sense when anyone can output anything
28:42 — Is an agency the best place to hone your craft right now?
32:22 — Juniors, remote work and rebuilding the intern-to-principal flywheel
34:11 — Hiring for fire in the belly, and why nobody reads case studies
35:54 — How to position yourself for your next move
36:47 — Honest feedback, a personal positioning statement and the Venn diagram
41:13 — Trust and shipping pace at the AI labs
42:42 — Advice for the labs: interfaces for infinite result sets
44:06 — Quality starts at the top: Airbnb and Brian Chesky
44:50 — Bring your most engaged users closer: Metalab and The Atlantic
47:24 — Engineering-led cultures and the pull to just build more
49:28 — Does this deserve to exist? Beatles or Stones, Netflix or Apple
51:18 — Quick fire: a product Sara loves, Doji
54:12 — After AGI: why Sara wants to be an art therapist
56:39 — Final thoughts, the Wise design event at the Tate, and why Europe keeps calling
Podden och tillhörande omslagsbild på den här sidan tillhör
Mostly Working. Innehållet i podden är skapat av Mostly Working och inte av,
eller tillsammans med, Poddtoppen.
