In this episode:

• Dia’s Morning Brief exists because most people stall at a blank text box, so the browser sends the prompt for you and opens a tab that briefs you on your day.
• Christine came back from two weeks’ holiday to a useful prototype that looked like ‘pure slop’, and rebuilt it around hundreds of hand-picked public-domain paintings from the National Gallery of Art.
• She designed Dia’s memory flows with lock icons everywhere, then watched user testing show that people were happy to tick the box.
• Arc shipped something new every week and ended up with around 35 branded features, whereas Dia builds and tests plenty and ships only what adds enough to the product.
• You get eight or nine nitpicks before an engineer has had enough of you, and Christine now ships those fixes herself the week after launch.
• A year ago she was asking her head of design whether she’d still have a job, and today she’s enjoying the work more than she has in years.

Chapters:

• 0:08 — Welcome to Mostly Working, and introducing Christine Røde
• 0:46 — From a Mostly Working meetup talk to Dia’s Morning Brief
• 1:21 — What Dia is, and how it has changed as AI has
• 4:08 — The blank text box problem, and a brief that arrives on its own
• 5:57 — The first prototype: beige, serifs and ‘pure slop’
• 7:00 — Finding the imagery: SVGs, big type or public-domain paintings
• 8:32 — Tom on what design adds to a prototype that already works
• 11:04 — Signal over volume: why the brief stays short
• 14:54 — Jonny on promising two or three things, and design reacting to engineering
• 16:37 — Why AI quality comes before design fidelity
• 19:01 — Onboarding, and persuading people to connect their tools
• 20:42 — From WhatsApp-era privacy to ‘you can all just read my email’
• 24:08 — Trust, and design’s part in earning it
• 26:45 — Designing Dia’s memory feature for users who weren’t worried
• 28:53 — What faster shipping does to the quality bar
• 30:09 — Why Dia ships less often than Arc did
• 32:01 — Designers shipping their own polish
• 34:12 — Working in seasons, and the hectic first month
• 38:24 — Custom tools for curating paintings and sketching interactions
• 42:08 — Real art as a brand decision, and hiring a painter
• 45:45 — What it means to be a designer now
• 48:17 — Quick-fire: free time after AGI, and Lime bikes
• 50:54 — The part of her craft Christine won’t hand to AI
• 52:09 — A rooftop garden, and the last 20% of vibe coding
• 53:31 — Final thoughts, and where to find Christine

Find us at mostlyworking.co

Click here to view the episode transcript.

In this episode:

• The Near Future consultancy is closing, and the podcast carries on every week as Mostly Working.
• We explain the three tests behind the decision: founder fit, how hard the work was to sell, and how much energy it gave us.
• At design + AI in Brighton, 30 senior designers built their own design tools in about an hour, from a Crazy Eights racing game to a skin analyser.
• The Mostly Working Summit runs at Monzo on 22 October: a free morning for 100 design leaders, then our final Build Your Own Design Tool workshop. Apply at mostlyworking.co.
• Tom walks through building the new brand, sketching the mark in Figma and having Claude Code generate palette tools, motion principles and even a jingle.

Chapters:

• 0:04 — The Near Future Podcast is becoming Mostly Working
• 1:03 — How the podcast started, 19 weeks in
• 2:17 — Winding down the Near Future consultancy
• 3:02 — Where the Mostly Working name comes from
• 4:30 — Founder fit, ease of sale and energy: why the consultancy - stopped
• 6:53 — Jonny on wanting to build, and joining a small tech company
• 11:09 — Tom's test: follow the most interesting conversations
• 13:21 — Thank you to Near Future clients
• 14:30 — What to expect from Mostly Working
• 17:24 — Build Your Own Design Tools at design + AI in Brighton
• 20:43 — Why the workshop builds keep getting better
• 23:15 — Hatch in Berlin: Paula Scher and Bob Baxley
• 26:33 — Announcing the Mostly Working Summit
• 27:57 — How the Mostly Working meetup grew from 70 to 160 sign-ups
• 31:45 — Summit details: Monzo, 22 October, fireside chats, roundtables and the workshop
• 36:13 — mostlyworking.co and the new brand
• 37:20 — Designing the M/W mark in Figma
• 39:09 — Claude Code generates palette tweakers, motion principles and a logo jingle
• 43:37 — Building brand tools by voice note from your phone
• 45:29 — Flora versus building your own image generators
• 47:22 — Tote bags, feedback and wrap-up

Mentioned in the episode

Mostly Working

• The Mostly Working Summit: AI x Design 2027 — Thursday 22 October 2026, Monzo, London
• Apply for the summit or book the workshop
• Mostly Working meetup

Events

• design + AI, Brighton, organised by Product Unleashed
• Hatch Conference, Berlin

People

• Tom Harman
• Jonny Burch
• Simon Amor, co-founder of Morse and co-host of the Mostly Working meetup
• Paula Scher, partner at Pentagram
• Bob Baxley, designer and former head of product design at Pinterest
• Domingo Widen, Fin
• Ini, Fin
• James Storer, Monzo

Companies

• Morse (formerly Sling Money)
• Monzo
• Fin
• Pentagram

Tools and products

• Claude Code
• Claude Opus
• Figma
• FLORA
• Suno
• OODA

About the pod

Mostly Working (formerly the Near Future Podcast) is a weekly conversation with design leaders about what they're building with AI. Send feedback to podcast@nearfuture.works, and find everything at mostlyworking.co.

Key takeaways

1. Dessn runs your codebase in the cloud so you can design on top of the real thing. No local setup and no coding, and the page you pull onto the canvas is your production homepage rather than a representation of it. Gabriella has waited her whole career to open the real homepage and work on it.
2. Read-only access to the repo was a focus decision before it was a security one. The first version of Dessn opened pull requests. They dropped it, because opening a PR isn’t the hard part anymore and it was never how they were going to win. Handoff now runs through MCP straight into Cursor or Claude Code.
3. Leaving the engineering gate where it is smooths adoption. Workflows are already changing, and change is heavy even when it’s powerful. Keeping one familiar checkpoint untouched means a design team only has to renegotiate the part that actually matters.
4. Dessn’s ideal customer profile turned out to be a culture, which is a very hard thing to qualify for. Every usual deterministic signal failed: vertical, size, stage. The proxy Gabriella watches for is whether a design engineer role already exists on the team.
5. An excited user counts for nothing if leadership isn’t bought in. Gabriella has watched a deal die at that gate with the user still cheering for it. Nim is the one who says no, and she’s learning to, because you only have so much time and where you spend it matters.
6. The handoff only ever existed because designers couldn’t reach the code. Remove the skill barrier and you remove the reason for the relay. Fewer people between the decision-maker and the executor gets you higher quality output and more speed.
7. Roles were built on skills, and skills are the part that’s collapsing. You knew Figma so you were a designer, you knew how to code so you were a dev. Gabriella never became a Figma expert and still considers herself a good designer, because the thinking layer was always the job.
8. Dessn’s learning curve is a curve of power. It’s mostly prompt-based, so there’s no tool to master, but there is a blank canvas problem: what can I even do in here? Ask it for a dropdown of every state a user could be in and it pulls them straight from your code.
9. The models caught up to the vision, and that’s what conviction buys you. It didn’t work, it didn’t work, it didn’t work, and then it worked. Dessn never rode the hype wave, because the product couldn’t deliver on the promises yet, and now it can.
10. The industry has gone from FOMO to fatigue after a slew of broken promises. Gabriella wants the headline to be exactly that: no broken promises. She’s blunt about competitors who answer the source-of-truth problem by creating another source of truth.
11. The designer’s job is heading towards curation, and that was always the high-leverage part. Deciding this is the right thing to do was always the leverage. Her worry, one year out, is how much Dessn builds for the requests arriving today versus the version of design work she thinks is coming.
12. “When did a feature become synonymous with a role?” An investor asked Gabriella where the visual editor was, given she builds for designers. Direct manipulation comes up constantly, and she thinks the industry should reopen that question rather than assume the answer.

Chapters

• 0:04 — Our first design tool founder: Gabriella Hachem and Dessn
• 1:18 — Montreal, first jobs, and never fitting the boxed roles
• 2:22 — She didn’t quit to build a design tool. She quit to solve a problem
• 3:34 — Design and prototype in your codebase, no local setup, no coding
• 6:42 — Jonny built a Temu version of Dessn for a hackathon
• 8:38 — Read access or write access to the repo, and why Dessn went read-only
• 11:04 — Opening a PR isn’t hard anymore, so that was never how they win
• 13:28 — Collapsing the talent stack: fewer people between decision and execution
• 14:51 — Roles were built on skills. She thinks they should be built on thinking
• 16:57 — Novelty budget, and a learning curve of power
• 19:36 — Pulling real production pages onto the canvas
• 20:48 — Every state defined in your code, and an automated Storybook
• 22:54 — When your ideal customer profile turns out to be culture
• 23:55 — Leadership is the real gate, and Nim is better at saying no
• 25:16 — Two years of founding: hard, fun, and a whole other level of hats
• 28:53 — How to know it’s your turn to take the risk
• 30:51 — Montreal to New York to London, and why not San Francisco
• 32:14 — Valio Con, a panel of design tool founders, and a week in SF
• 33:11 — Frenemies, and the one thing nobody else does
• 35:06 — Form an opinion about how you want to work, then go work that way
• 40:03 — Creating another source of truth makes the problem worse
• 41:57 — From FOMO to fatigue, and the models catching up to the vision
• 43:30 — Speed, latency, and 120 frames per second
• 46:17 — Proactive agents, direct manipulation, and the designer as curator
• 49:33 — “When did a feature become synonymous with a role?”
• 50:38 — Personalised workflows, plugins, and Dessn as infrastructure
• 52:57 — React Native now, iOS Native and a Mac Mini next
• 54:23 — Quick fire: a wine bar, a spatula, deleting X, and watching your mum use ChatGPT
• 59:19 — Where to find Gabriella, and the infinite canvas beta

Mentioned in the episode

Gabriella and Dessn

• Gabriella Hachem — https://www.linkedin.com/in/gabriella-hachem/
• Gabriella on X — https://x.com/gabriellahach
• Dessn — https://www.dessn.ai
• Nim Cheema, co-founder and CTO — https://ca.linkedin.com/in/nim-cheema
• ‘Collapsing the talent stack’, Gabriella’s piece on Scott Belsky’s essay — https://medium.com/@gabriella_46676/collapsing-the-talent-stack-40eb849a2f97
• Dessn’s $6m seed round, led by Connect Ventures with Betaworks and N49P — https://techcrunch.com/2026/05/12/dessn-raises-6m-for-its-production-focused-design-tool/
• Kevin Grajeda, Dessn’s employee number one, who Gabriella met on X — https://www.linkedin.com/in/kevingrajeda/

Companies and past roles

• Planned, where Gabriella was head of product and Nim was head of AI — https://planned.com
• Unsplash, where Nim worked — https://unsplash.com
• Betaworks, who hosted them in New York — https://www.betaworks.com
• Connect Ventures, who led the seed — https://www.connectventures.co

People

• Scott Belsky, author of ‘Collapsing the Talent Stack’ — https://www.scottbelsky.com
• Stephen Haney, foun...

Key takeaways

1. Build the tool while you build the work. Frank made his design tool feature by feature during his first project at Anthropic, adding each part the moment he hit the need for it. A share function when he wanted to show his PM, then a Loom-style recording feature the next day, camera feed and all, in about three prompts.
2. Load a control and a variant side by side and click them in sync. The most unusual thing he built steps two prototypes through the same screens on a single click. He is hoping Figma steals the idea.
3. Most people inside a company have never seen the screens they own. Everyone at Headspace got the app free, so nobody had laid eyes on an upsell screen in years. A live side-by-side of current state and proposed state is what gave leadership the confidence to green-light the work.
4. Begin anywhere. Frank has turned John Cage’s line into a working design principle. What matters is having something to click, see and feel, and the details get figured out along the way.
5. Variety compounds differently to mastery. Energy drinks at Red Bull, meditation at Headspace, then social media at Instagram: every jump was chosen on purpose. He calls himself a Swiss Army knife of design, and reckons it is what makes someone stand out in an interview panel.
6. Learning the tools on the company dime is a real benefit worth asking for. DoorDash set a goal it called Design 2028, where every designer becomes a design engineer, and gave people the time and space to get there. Frank admits he complained about it at first and is now grateful he did not have to learn it in his own hours.
7. A lot of people working hard on something is a reason for empathy, and the metric still decides whether it worked. Frank got pushback from a former colleague for publicly critiquing a feature. His read: designers who want to talk business have to be able to talk about failures like adults, because there is not a lot of emotion in business.
8. Do the next project’s pre-work before you get to it. Frank queues up a foundation on the Tuesday so that Monday’s project starts from something rather than a blank page. Half of it may be throwaway, and a PM running agents can now come back in a morning with what used to take a two-week sprint.
9. Solve one of your own problems and it becomes the standout project in your portfolio. For anyone in an organisation with no AI budget and no guidance from above, that is Frank’s route through. A grocery list app or a plugin for a digital guitar amp both count, imperfect included.
10. Old, slightly bad technology is a good counterweight to this pace. Frank keeps a five-inch black-and-white Quasar countertop television on his desk, one channel and a radio dial. He calls it poorly designed and perfectly designed at the same time, the same pull that took people back to vinyl.
11. Close the 100 tabs and learn one thing today. His answer to feeling overwhelmed in an AI-first company. Nobody has landed on a perfect process yet, which means the process is still yours to shape, and he found that anxiety and excitement flipped over into each other almost overnight.

Chapters

• 0:04 — Bureau, Edenspiekermann, Headspace, Instagram, DoorDash: the run-up to Anthropic
• 2:18 — A day at Anthropic, where a 12-month roadmap makes no sense
• 4:32 — The LinkedIn post that started all this
• 5:03 — A quiet onboarding week and an enterprise Claude account
• 7:46 — “How do I share this?” A share function, then a Loom feature, in three prompts
• 9:14 — Shortcut keys, screen glimmers and museum wallpapers
• 10:18 — Accidentally rebuilding Claude Design, backwards
• 11:24 — Sunshine Shop, WooCommerce and total cowboy mode
• 12:33 — Will the Frank tool still be around in 12 months?
• 13:37 — Two prototypes side by side, clicking in sync
• 15:41 — Where the line sits between human and AI
• 16:43 — Mega tools, workspaces and a live command line
• 18:54 — Energy drinks to meditation to social media: the Swiss Army knife career
• 21:07 — Mastery versus invention, and naming the Frank Mega Tool
• 23:52 — DoorDash’s Design 2028, and a quarterly meetup as a forcing function
• 27:30 — Accountants, lawyers, and doing your taxes with an AI tool
• 29:19 — The weirdest design processes he has come across
• 32:19 — John Cage: begin anywhere
• 34:13 — Critiquing an old colleague’s feature, and the fallout
• 37:00 — Advice for designers with no budget and no backing
• 41:14 — Mathematicians, art, Spotify and Suno
• 42:55 — Quickfire: AGI takes over, what do you do with the time?
• 43:59 — Quickfire: the five-inch Quasar TV that is perfectly badly designed
• 45:59 — Staying grounded when the pace is this intense
• 48:41 — Close your 100 tabs and learn one thing today

Mentioned in the episode

Frank

• Frank Bach’s website — francoisbach.com
• Frank on LinkedIn — linkedin.com/in/francoisbach
• Frank on X — x.com/zendadddy
• Sunshine Shop, his creative collective and e-commerce store — sunshineshop.la
• LA Design & Dev, the Los Angeles meetup he organises — ladesignanddev.com

Companies and past roles

• Anthropic — anthropic.com
• Bureau, the design studio Frank co-founded and later closed
• Edenspiekermann, where Red Bull was the main client — edenspiekermann.com
• Red Bull — redbull.com
• Headspace — headspace.com
• Instagram — instagram.com
• DoorDash — doordash.com

People

• Jenny Wen, Head of Design at Claude, whose Hatch Conference talk Jonny saw — jennywen.ca
• Andy Puddicombe, the English co-founder of Headspace — headspace.com/content/video/meet-andy-puddicombe/923
• John Cage, composer, source of “Begin anywhere” — johncage.org

Tools and products discussed

• Claude — claude.ai
• Claude Design — anthropic.com/news/claude-design-anthropic-labs
• Figma — figma.com
• Cursor — cursor.com
• Loom — loom.com
• Sketch — sketch.com
• Adobe Illustrator and Photoshop — adobe.com
• WordPress — wordpress.org
• WooCommerce — woocommerce.com
• Dropbox — dropbox.com
• GitHub — github.c...

Key takeaways

1. Being deliberate about what you give away is most of the job now. Every August guest drew the line somewhere different, and every one of them drew it. Jonny’s version: even when it is slower, and even when an AI would do it better, make sure the decisions being made are still yours to make.
2. Make one decision with an agent and fifty get made for you. They arrive interlinked, so working out which ones landed and why costs more than piling on another prompt. You end up with work that is technically yours and nothing in your head connecting it.
3. A learning loop is an agent that remembers its own mistakes. Matt Webb’s Poem/1 fulfilment agent handled queries from 700 units by turning each new exception into a standing rule. Jonny’s caveat is that the abstract version convinces nobody, which is why “everyone should have agents” lands so flat.
4. Agents accumulate cruft, so prune them. Jonny moved his main agent from NanoClaw version one to version two this week and deleted about half the notes it had written to itself over six months. What survived reads as concise instructions instead of six months of contradictions.
5. When everyone can make the impressive thing, impressive moves. GPT-6 Astra put one-shot 3D and cloned games within reach of anyone with a prompt. Jonny’s yardstick for where the value went: his dad, a vicar, could now get further on a demo video than Jonny could have managed five years ago.
6. Understand the grain of the medium you are designing against. Randy Hunt’s framing was the old-school graphic designer collaborating with the printer to get intent onto the page. Having an agent write your code is a different skill from understanding what that code does to your design.
7. AI writing feels thin because the thinking happened somewhere else. Matt Webb can feed a model everything he has ever published and still fail to stand behind what comes back. It may convince everyone else, and it is still not something you can point at and call yours.
8. Build the silly project yourself or you will never learn the tool. Demos online tell you nothing about where the limits sit. Tom is close to downloading Blender for exactly this reason, and Jonny built a full 3D Labyrinth in two hours partly to find out what two hours now buys.
9. The impressionists are the closest analogue to this moment. Photography arrived, exact landscapes stopped being the point, and painters who already had mastery chose what to break. Whoever does best next will understand the technology underneath well enough to bend it.
10. Skills get better when you rewrite their output by hand. Tom burned a month of Claude Fable 5.1 credits in half an hour, then rewrote both research skills’ results manually and fed those back as the template. A 10,000-word essay he would never read became a 400-word list of links.
11. The craft question matters more than the job title. Tim, a listener, asked whether the show could branch out beyond senior design folks. Yes, with a lens: creative technologists, engineers, product people and founders are all fair game, as long as the question stays how AI is shifting the role and craft of design.

Chapters

• 0:04 — Intro: back to school, and a cosy September episode
• 0:59 — Rounding up August’s four guests
• 1:38 — James Storer: Monzo’s fake app and its biggest token spender
• 3:29 — Randy Hunt: Notion, SVA, and two full-time roles
• 4:36 — The printing press, and the grain of the medium
• 6:30 — Patrick Lewis: four times a founding designer
• 7:57 — T-Rex arms, and shaping a role around what you enjoy
• 8:51 — Matt Webb and an inventor’s practice
• 9:49 — Poem/1’s self-learning fulfilment loop
• 11:31 — The Gmail triage loop Tom hasn’t built yet
• 12:43 — OpenClaw, and agents that rewrite their own memory
• 13:36 — Lobotomising an agent onto NanoClaw version two
• 16:59 — Where every guest drew their own line
• 17:22 — Are we losing our leverage to agents?
• 18:30 — GPT-6 Astra, and what happens when everyone can do it
• 22:20 — Make one decision, get fifty
• 24:10 — Is design an art or a science?
• 26:47 — Inventors, not technologists
• 28:44 — Why you have to build the silly project yourself
• 30:20 — The impressionists, photography, and our toddler phase
• 31:50 — Listener feedback: branch out beyond senior design folks?
• 33:45 — Grading the podcast, and 36 YouTube subscribers
• 36:29 — Design at the core, but a wider cast
• 39:40 — Labyrinth, one-shotted in two hours
• 42:45 — Burning a month of Claude Fable credits on skills
• 45:40 — Consolidating everything into one Dia browser
• 47:39 — Design leadership circles
• 49:12 — Hatch Berlin, Brighton, and save the date for 22 October
• 53:16 — Wrap up

Mentioned in the episode

The August episodes we round up

• Ep #12: James Storer (Monzo) — share.transistor.fm/s/ab6c1806
• Ep #13: Randy Hunt (Notion & SVA) — share.transistor.fm/s/29c8593b
• Ep #14: Patrick Lewis (Klay) — share.transistor.fm/s/8d030297
• Ep #15: Matt Webb (Inanimate) — share.transistor.fm/s/c39062f5

People

• James Storer, Senior Product Design Lead at Monzo — linkedin.com/in/james-storer-5b543625
• Randy Hunt, Head of Design at Notion and chair of MFA Design at SVA — randyjhunt.com
• Patrick Lewis, founding designer at Klay — patricknlewis.com
• Matt Webb, writing at Interconnected — interconnected.org
• Christine Røde, product designer at The Browser Company — linkedin.com/in/christinerode
• Lenny Rachitsky, host of Lenny’s Podcast — lennysnewsletter.com/podcast
• Tim, the listener who wrote in with the feedback

Companies and past roles

• Monzo — monzo.com
• Notion — notion.com
• SVA MFA Design — design.sva.edu
• Klay — klay.vision
• Inanimate, Matt Webb’s current company — inanimate.tech
• The Browser Company — thebrowser.company

Tools, models and products discussed

• GPT-6 Astra — openai.com/index/gpt-6-astra
• Claude Fable 5.1 — anthropic.com/claude/fable
• OpenClaw — openclaw.ai
• NanoClaw — nanoclaw.dev
• Grok Bot — x.ai
• Cursor — cursor.com
• Dia, the browser by The Browser Company — diabrowser.com
• Blender — blender.org
• Poem/1, Matt Webb’s poetry clock — poem.town
• Labyrinth, the Ravensburger board game — ravensburger.co.uk

Jonny’s own projects

• Labyrinth online, built in two hours — labyrinth.jonny.ooda.run
• Ooda, secure collaborative hosting for teams — ooda.run
• designtools.fyi, run by an agent that maintains itself — designtools.fyi

Events

• Mostly Working, the London design mornings Tom runs — lu.ma/mostly-working
• Hatch, Berlin, 16–18 September, where Jonny is speaking on the 18th — hatchconference.com
• design + AI, Brighton, 24–25 September, where Tom and Jonny run a workshop — designplusaisummit.com

About the pod

Key takeaways

1. Shape your role around your practice. Matt focuses on human-computer interaction, small groups and the physical world, rather than chasing a job title.
2. AI is a rubber duck, not a co-writer. Its value is helping you think aloud without replacing your voice.
3. Universal Basic Agency. AI can help non-experts cross the single technical hurdle blocking an idea.
4. Agents can turn exceptions into policy. Poem/1's fulfilment agent proposes reusable rules for a shared SOP, so problems are solved once.
5. Reports can reveal more than dashboards. Daily narratives surface patterns you did not know to track.
6. Make deliberate time to learn new tools. Matt blocks at least a day a week because the gains compound.
7. Building on a model means borrowed time. Poem/1 relies on a retiring model whose strange tone newer models cannot reproduce.
8. Poem/1 is an invisible social network. All 700 clocks show the same poem at the same moment worldwide.

Chapters

• 0:04 — Intro and Matt's career throughline
• 5:56 — Leaving leadership to build
• 7:28 — AI and unblocking imagination
• 13:02 — Keeping AI out of your voice
• 17:35 — ChatGPT and tldraw as a rubber duck
• 24:12 — The imagination bottleneck
• 29:04 — Universal Basic Agency
• 34:38 — The story behind Poem/1
• 44:44 — A retro AI product and invisible network
• 52:08 — Agentic Kickstarter fulfilment
• 1:02:54 — Reports versus dashboards
• 1:06:40 — Quickfire questions

Mentioned in the episode

Matt

• Matt Webb's blog, Interconnected — interconnected.org
• Unoffice Hours (his 30-minute call booking system) — interconnected.org/home/2020/09/24/unoffice_hours
• Acts Not Facts, his consulting practice — actsnotfacts.com

Companies and past roles

• Berg — the design studio Matt co-founded with Jack Schulze and Matt Jones
• Google Research
• R/GA Ventures
• Inanimate — Matt's current startup — inanimate.tech, newsletter at news.inanimate.tech
• Betaworks — betaworks.com

People

• Jack Schulze — Berg co-founder
• Matt Jones — Berg co-founder
• Daniel Fogg — Matt's co-founder at Inanimate
• Alex Hulme — founder and creative director of Approach, the industrial design agency that helped bring Poem/1 to market — approach.studio, Poem/1 case study at approach.studio/work/poem1
• Tom Armitage — helped with Poem/1's firmware — tomarmitage.com

Tools and products discussed

• ChatGPT — chatgpt.com
• Claude Code — claude.com/product/claude-code
• tldraw — tldraw.com
• BackerKit — backerkit.com
• Floship — floship.com
• M5Stick, made by M5Stack — m5stack.com
• Friend, the AI pendant — friend.com

Matt's own projects, mentioned

• Little Printer, made at Berg — berglondon.com/projects/bergcloud
• Poem/1, Matt's poetry clock — poem.town
• Galactic Compass, the app that points to the centre of the galaxy — apps.apple.com/gb/app/galactic-compass
• A 270kg book vending machine that tweeted when it sold a book

About the pod

• Podcast feedback: podcast@nearfuture.works
• Find us at https://nearfuture.works

Chapters
0:04 — Meet Patrick Lewis, founding designer at Klay
1:05 — Building Clay: an AI music app made by musicians, not sales bros
2:27 — Content still has to resonate with humans, AI-generated or not
6:27 — Consistency builds trust: designing believable AI video avatars
10:16 — Figma bulls and bears: Jonny's rebuy vs. Patrick's near-total exit
12:37 — Where Figma still earns its place: branding and early sketches
13:42 — Founding designer for the fourth time: getting context fast with AI
16:51 — Building a pun-named playground to prototype without baggage
19:17 — Layoffs, flattened experience curves, and adopt-or-die
27:09 — Knowing yourself, saying no, and the founding designer's tightrope
36:44 — Capacity has grown, but so has the pressure to do it all
39:23 — Quickfire: chairs over AGI, decision-making he won't outsource, and granola

Patrick Lewis — https://patrickjlewis.com

Tools and products discussed
Figma — https://www.figma.com
Claude — https://claude.ai
ChatGPT — https://chatgpt.com
GitHub — https://github.com
Kit (newsletter platform) — https://kit.com
Granola — https://www.granola.ai

Find us at nearfuture.works

Chapters

• 0:04 — Two institutions, one calendar: how Randy keeps both plates spinning
• 3:00 — Agents that absorb the busy work so he can stay close to the craft
• 5:29 — Dogfooding vs. knowing the competition: you have to do the work twice
• 8:22 — One place for everything, or a walled garden? Jonny re-buys Notion
• 11:10 — Battle-tested vs. multi-model: why living in one suite stops working
• 13:22 — Good, fast, cheap: Randy refuses to pick two
• 15:17 — Quality isn't the screen being right, it's how many people hit it wrong
• 19:05 — The handoff is a telephone game: why passing the baton kills the work
• 22:52 — Designers were shipping production code before the coding agents arrived
• 25:42 — He built it in a weekend. The PR took two weeks to review
• 29:35 — Writing code got cheap. Writing good code didn't
• 30:46 — Emergent chat vs. written-down skills: making the system legible to everyone
• 33:07 — Working in the open so you can see how your colleagues actually prompt
• 35:42 — Build it yourself or wait for someone to sell it: micro tool making
• 39:04 — The barbell curriculum: what endures vs. what has to stay malleable
• 43:20 — Directing a photographer vs. directing a model: same skill, different studio
• 44:57 — Where's the punk rock student who refuses to use AI?
• 47:58 — He went home for the summer and built an ERP for his dad's warehouse
• 49:33 — Quick fire: AGI takes over, so what do you actually do with the time?
• 51:37 — Wrap up, and a promise to come back for the music conversation

Mentioned in the episode

Randy

• Randy J. Hunt — https://www.randyjhunt.com
• Randy on LinkedIn — https://www.linkedin.com/in/randyjhunt/
• Product Design for the Web (New Riders, 2013) — https://www.randyjhunt.com
• SVA MFA Design, where Randy is chair — https://design.sva.edu
• Randy speaking at Hatch Conference, Berlin, September — https://www.hatchconference.com/speakers/randy-hunt

Companies and past roles

• Notion — https://www.notion.com
• Notion Agents, his current favourite app — https://www.notion.com/product/agents
• Etsy — https://www.etsy.com
• Artsy — https://www.artsy.net
• Grab — https://www.grab.com
• Deliveroo — https://deliveroo.co.uk
• Uber Eats — https://www.ubereats.com

Tools and products discussed

• Claude — https://claude.ai
• Claude Code — https://claude.com/product/claude-code
• Claude Cowork — https://www.anthropic.com/product/claude-cowork
• MCP (Model Context Protocol) — https://modelcontextprotocol.io
• ChatGPT — https://chatgpt.com
• Google Gemini — https://gemini.google.com
• Google Workspace / Google Sheets — https://workspace.google.com
• Slack — https://slack.com
• Figma — https://www.figma.com
• Weavy, acquired by Figma and relaunched as Figma Weave — https://www.figma.com/blog/welcome-weavy-to-figma/
• Paper — https://paper.design
• Pencil — https://pen.dev

Events

• Hatch Conference, Berlin, September — https://www.hatchconference.com

CHAPTERS

• 00:00 Introducing James Storer
• 02:12 Why go back to doing the work after leading teams
• 10:13 The founder's fight: ship it, or make it better
• 14:47 Being a loud mouth turned out to be the skill
• 18:35 AI lowers the floor, so who raises the ceiling
• 25:30 Everyone an expert in everything, or still not enough hours
• 28:01 Building hardware when you can't solder
• 30:07 Should there be an app store with one user
• 33:43 Is the value the software, or making it with your kid
• 36:08 The UNO clone his six-year-old wanted to explode
• 40:01 A 69-year-old vicar and the church website
• 41:31 The messy middle: AI inside a regulated bank
• 43:44 A fake Monzo with fake money
• 49:36 Handoff should be a handshake
• 54:16 "What will you do to make me open you every day?"
• 60:39 Quickfire
• 68:31 Wrap

PEOPLE

• James Storer — Design Lead, AI & Personalisation, Monzo https://www.linkedin.com/in/jamesstorer/
• Avi Ashkenazi, Deel — previous Near Future guest, on raising the floor and the ceiling https://www.linkedin.com/in/aviashkenazi/
• Jake Knapp — author of Sprint — https://www.jakeknapp.com
• Simon from Morse — spoke alongside James at Mostly Working https://www.linkedin.com/in/simon-amor/

COMPANIES

• Monzo — https://monzo.com
• Deliveroo — https://deliveroo.co.uk
• Freetrade — https://freetrade.io
• Habito — https://www.habito.com
• Klarna — https://www.klarna.com
• Twist — James's design agency
• Jitty — the house-buying startup James co-founded

TOOLS AND PRODUCTS MENTIONED

• Cursor — https://cursor.com
• Claude Code — https://claude.com/product/claude-code
• ChatGPT — https://chatgpt.com
• Figma — https://figma.com
• Figma Make, Figma Weave, code layers — announced at Config
• FigJam — https://figma.com/figjam
• Lovable — https://lovable.dev
• Stitch — https://stitch.withgoogle.com
• NotebookLM — https://notebooklm.google.com
• Notion — https://notion.com
• Linear — https://linear.app
• Abstract — version control for Sketch files
• Sketch — https://sketch.com
• Three.js — https://threejs.org
• React Native — https://reactnative.dev
• Nano Banana — Google's image model, via Gemini
• Hex — open-source dictation app
• Wispr Flow — https://wisprflow.ai
• Wabi — personal app builder for iPhone
• MeatLayer — agents paying humans for real-world tasks — https://meatlayer.ai
• M5Stack — small dev boards — https://m5stack.com
• Lumi / ROLI — light-up keyboards, now open sourced — https://roli.com
• BigQuery — https://cloud.google.com/bigquery

ALSO REFERENCED

• Config 2026 — Figma's conference — https://config.figma.com
• Mostly Working — Tom's design and AI meetup in London https://luma.com/mostly-working
• Sprint, by Jake Knapp, John Zeratsky and Braden Kowitz
• UNO — the card game James and his son cloned as "Duo"

NEAR FUTURE

• Podcast — https://nearfuture.works/podcast
• Get in touch — podcast@nearfuture.works

They reflect on what the recent guest run has taught them, dig into Backyard Designers Club and the hooha it sparked, and talk through why design and engineering leadership is such a hard place to be right now. 

Plus: Tom built a 3D chess game for his 5-year-old using Opus 5 and Three.js, and Jonny's Ooda gets a quiet YC moment.

Chapters

0:04 — Intro: Jonny's back (and has internet again)

3:44 — Reflections on a month of guest episodes

8:20 — The common thread: leaders who build

17:00 — AI raises the floor, not the ceiling — but the ceiling hasn't moved

21:35 — The AI slop problem: Hacker News, Substack's detector & LinkedIn's report button

22:08 — Ellipsis and the proof-of-work problem for human writing

26:10 — Ini's wireless-in-the-browser project: the opposite of slop

21:32 — Backyard Designers Club: craft, sound design and game-like delight

29:13 — How fast an original idea can be cloned

31:21 — The hoo-ha: exclusion, community claims, and what a community actually is

38:53 — Who's on Backyard Designers — and what it says about design's visible/invisible divide

41:35 — Engineering leadership burnout: Gergely Orosz's three reasons leaders are quitting

44:35 — Is design leadership 12 months behind engineering on the same trajectory?

51:57 — Tom's AI experiments: agent skill pack + a 3D underwater chess game

1:00:01 — Jonny's Ooda project gets a YC validation moment

1:00:42 — What's coming up: Build workshop, Wednesday event with Christine Røde (The Browser Company), Brighton in September

Useful Links

Backyard Designers Club — https://backyarddesigners.club

Broke Ass Designers — the commentary/reaction version - https://brokeass-designers.fly.dev/

Ellipsus — proof-of-work writing tool https://ellipsus.com/

Gergely Orosz (The Pragmatic Engineer) — engineering leadership trends post https://www.linkedin.com/posts/gergelyorosz_interesting-recent-trend-im-observing-share-7487619068187852800-ceQ2/

Ooda — Jonny's internal tool hosting project — https://ooda.run

Podcast feedback: podcast@nearfuture.works

Near Future Build Workshop (Aug 13th): https://nearfuture.works/build

Find us at https://nearfuture.works

• (00:00) - Intro & Guest Welcome
• (00:52) - Joining AirOps: The 90-Day Plan
• (05:13) - Building the AI Content Design Skill
• (12:38) - Product Design vs. Design Operations
• (20:39) - Neurodivergence, Communication & Knowledge Systems
• (32:04) - Daily Reflection Practices
• (36:22) - Palmwine Wireless: Building a Nigerian Radio
• (43:41) - Sound Design Deep Dive
• (49:26) - Building the Music Library with Suno
• (52:45) - Quick-Fire Questions
• (59:00) - Outro & Thanks

- [Domingo Widen](https://www.linkedin.com/in/domingowiden/)
- [Monologue by Every](https://www.monologue.to)
- [Andrej Karpathy](https://x.com/karpathy)
- [Wes Kao](https://www.weskao.com)
- [Lenny’s Data](https://www.lennysdata.com/)

## More about Ini

- [Ini’s personal website](https://iniabiodun.com/)
- [Palmwine Wireless](https://radio.iniabiodun.com/)
- [Ini on LinkedIn](https://www.linkedin.com/in/inioluwa-abiodun/)

Find us at nearfuture.works

Chapters
0:04 — Intro & guest welcome
0:49 — What day-to-day design at Wise looks like right now with AI
2:27 — The early prototyping kit that made AI feel "magical"
4:45 — Solo missions vs. collaboration in AI-driven work
6:07 — The "agent-operator" workflow trap
7:16 — Why scale (8,000 employees) changes the risk calculus
7:33 — How AI adoption varies across design disciplines
8:58 — The unglamorous ops work behind AI prototypes (InfoSec, workflows)
16:49 — Talent development: grads, internships, apprenticeships
19:03 — The three traits that matter most: smart, ambitious, curious
19:49 — Why design disciplines aren't collapsing
20:41 — "AI raises the floor, not the ceiling" — differentiation matters more
24:32 — Josh's 2x2 framework: when to let AI/PMs skip design entirely
29:37 — AI as an "averaging machine" and the tells of AI-generated work
32:13 — A WALL-E-inspired vision of the future
33:15 — Betting on human-curated, handcrafted experiences
34:36 — Josh's secret AI-built app project
35:15 — Josh's actual design process, tool by tool
37:27 — His AI agent fleet: Claude Code, Cowork, and more
39:47 — Bounded validity and lessons from engineering-first principles
42:24 — Quick-fire round begins
43:16 — What they'll never hand over to AI: curation
44:22 — The last tool they deleted (and why)
45:25 — What's inspiring them right now
46:58 — Wrap-up & where to find Hollie and Josh

Useful Links

Josh Payton, VP of Design at Wise
https://www.joshpayton.com/
https://www.linkedin.com/in/joshpayton

Hollie Lubbock, Design Director at Wise
https://www.hollielubbock.com/
LinkedIn: https://www.linkedin.com/in/hollielubbock/

Podcast feedback: podcast@nearfuture.works
Near Future workshop (Aug 13th): nearfuture.works/build
Find us at nearfuture.works

Click here to view the episode transcript.

Domingo shares how designers at Fin have moved from prototyping in sandboxes to shipping directly to production, and unpacks his team's mission to build the platform and tools that let every team move from concept to production faster with AI.

The conversation traces the story of getting designers into developer environments — from a top-down mandate and painful early setup sessions to a smoother, engineer-assisted process — and how that groundwork now lets designers self-serve with Claude. 

Domingo introduces "Surge Intelligence," an encoded knowledge base that powers design-system skills so AI can enforce front-end conventions without a human policing every decision, and describes the PR-style review process for designers who want to contribute their own skills.

The back half explores where design is headed: the enduring value of human judgment and "taste" built through reps, the edge cases only builders learn from experience, and why Domingo believes AI will keep design work capped around 85% for a while. They close with quick-fire questions on tools he wishes existed (a home for aligning on live prototypes, since Figma hand-off has broken down), his favourite product (MyMind), and what he'd do with his free time in an AGI future.

Chapters

0:00 – Intro & welcoming Domingo Widen

0:52 – How the Front-End Infrastructure team works today

2:20 – Is Domingo still a "product designer"?

3:49 – Splitting time between infra, design system, and product work

4:22 – Getting designers into developer environments: the origin story

9:45 – Optimizing the setup process with engineering

14:20 – From constant support requests to a self-serve steady state

11:29 – Claude as a fallback for technical hiccups

12:14 – Enforcing design consistency without slowing the company down

14:02 – Introducing Surge Intelligence and how design skills read from it

15:57 – Serving both designers and engineers with shared knowledge

18:13 – Do designers see AI codifying their instincts as a threat?

20:09 – The 85% problem: where human judgment still wins

23:52 – Codifying skills: the PR-and-review process for Team 2X

25:11 – Human judgment vs. automated skill approval

26:15 – Predictions for how the designer role evolves industry-wide

28:20 – Foundational design knowledge that AI won't replace

30:37 – On "taste," reps, and gut instinct

32:38 – The superpower (and blind spots) of designing in code

34:24 – Is Claude enough, or will tools fragment further?

36:20 – The missing tool: aligning on live prototypes instead of Figma

39:28 – Zooming out: is this a temporary infra push or years-long work?

41:12 – Why this becomes "training the brain" rather than a one-off project

42:56 – Optimizing for designers, engineers, and PMs alike

43:56 – Quick-fire: coffee, drawing, and life after AGI

45:37 – Quick-fire: the product Domingo wishes he'd designed (MyMind)

Useful Links

Domingo’s website: https://www.domingowiden.com/

Domingo on Linkedin: https://linkedin.com/in/domingowiden

Mymind: https://mymind.com/

Podcast feedback: podcast@nearfuture.works

Near Future workshop (Aug 13th): nearfuture.works/build

Find us at nearfuture.works

Click here to view the episode transcript.

The conversation dives deep into Deel's AI-forward culture: designers losing Figma seats after 30 days of disuse, "super users" burning through massive AI token budgets, twice-weekly AI showcase sessions, and how much of the design workflow has shifted from Figma into tools like Claude Code. Avi also discusses building internal tools like Links (an Obsidian-style knowledge brain) and an AI-powered portfolio screening tool for the ~300 daily design applications Deel receives.

The back half covers hiring in the AI era — including why uniqueness and craft still matter more than ever, the tension between AI making collaboration seamless versus more isolating, and Avi's "spicy take" that no one should manage designers without being a staff-level practitioner themselves. They close on how AI-freed time should be spent: hunting for the next big cross-functional bet rather than busywork.

Chapters
0:00 – Intro & welcoming Avi Ashkenazi
0:33 – Building Vidit: Deel's in-house Screen Studio alternative
3:49 – 600 daily users and open-sourcing (for free) at viddit.video
7:56 – How AI tools have changed ways of working at Deel
10:25 – Deel's build culture, one-click dev onboarding, and losing Figma seats
14:44 – Where the design industry is heading with AI (Shopify vs. Deel)
18:44 – Token budgets, "super users," and adoption curves across the team
23:54 – Scaling judgment: building "Links" and "Launchpad" brains
35:23 – AI-assisted hiring: screening 300+ design applications a day
43:28 – What AI gets wrong when judging portfolios (uniqueness & self-affirmation)
50:18 – Where the freed-up time should go & Avi's spiciest take on AI and craft
57:04 – Wrap-up & useful links

Useful Links
Vidit (Avi's screen recording tool): viddit.video
Avi's personal site: superavi.com
Deel careers: deel.com/careers

Podcast feedback: podcast@nearfuture.works
Near Future workshop (Aug 13th): nearfuture.works/build
Find us at nearfuture.works

Click here to view the episode transcript.

They reflect on how the meetup format has evolved as more people have hands-on AI experience, shifting from first experiments to practical, scaled adoption inside teams, with more focus on token costs and measurable value.

They discuss Figma Config releases (Motion, Figma Weave, Sites/Make, and upcoming code components), debating canvas-first vs code-first workflows and the “source of truth” question for enterprise teams.

Jonny summarizes a Near Future Works post on AI transformation: set a coherent strategy, secure org-wide buy-in, and maintain craft/quality so AI speed doesn’t produce worse outcomes. They also touch on music ownership, a Claude Code World Cup calendar prompt, Anthropic “Fable” returning and Sonnet 5, Z.ai’s open-weights GLM model, a keyboard-shortcuts tool, and an open online workshop (Aug 13) to build design tools.

00:00 Welcome and agenda
00:12 Mostly Working recap
00:52 Podcast updates and banter
03:01 Meetup lessons so far
07:13 Figma Config reactions
12:39 Source of truth debate
22:51 Jagged frontier and craft
28:11 Quality Bar Cannot Drop
28:47 Barbell Strategy for AI
30:18 Craft Beyond Output
32:37 Atrophy and Writing Muscle
35:47 Craft Across Disciplines
36:40 Music Ownership Stories
41:22 Claude Calendar World Cup
42:57 AI News Frontier Models
45:27 Beautiful Shortcuts Tool
46:57 Open Workshop Invite
49:48 Signup Feedback and Wrap

Links:
Mostly Working: https://luma.com/mostly-working
Near Future blog post: https://nearfuture.works/updates/a-new-jagged-frontier-why-craft-still-matters
Figma's config keynote: https://config.figma.com/san-francisco/session/829e6ced-3257-4f5c-b675-aa72f4d1f98f/
GLM5.2 by z.ai: https://z.ai/
https://shrtcts.click/
Our next workshop: https://nearfuture.works/build

As always, feedback welcome at podcast@nearfuture.works

Find us at nearfuture.works

Click here to view the episode transcript.

Jonny starts by recapping his golf trip to Mallorca — where a non-technical friend used Lovable to vibe-code a fully functional scoring app for their annual 16-man Ryder Cup-style tournament, the Stryder Cup. It's a perfect case study in personal software.

From there, Tom brings a meaty topic: the "death of the middle manager" narrative gaining steam in the press. Drawing on his five years as Director of Product Design at Monzo, Tom and Jonny debate whether this is a real structural shift or a post-ZIRP overcorrection — and what managers should actually do to stay relevant.

In the quick-fire round, Jonny introduces Ideogram, a small but opinionated image model that excels at visual design and can be self-hosted and fine-tuned on your own work. Tom follows up with Midjourney's unexpected pivot into health and wellness medical devices — and what it says about where AI companies might be headed.

They close by testing themselves on "In the Weights" — a site that checks if your name appears in AI training data. Jonny is in (top 25% British product designer). Tom is not.

Finally, they tease an upcoming half-day AI design workshop open to individuals — coming late July/early August. Email podcast@nearfuture.works with feedback, ideas, or complaints.

Chapters
0:00 — Welcome & podcast updates
New email address: podcast@nearfuture.works

1:23 — Guests & format chat
Guests are coming — but this pod isn't going guest-only

2:00 — Jonny's golf trip to Mallorca
16-man Ryder Cup-style tournament, Mallorca, 35-degree heat

4:06 — The Strider Cup app — vibe coded on Lovable
How a non-technical friend built a live scoring app with AI — and why personal software is the perfect use case

8:59 — Can AI improve your golf game?
Spoiler: Jonny is a lost cause

10:49 — Are middle managers finished?
Tom introduces the "Middle Managers in the Firing Line" article and his Monzo perspective

18:36 — Tom's take: the functions matter, not the title
Is Jack Dorsey right? Tom's nuanced answer on direction, psychological safety, and player-coaches

22:02 — Jonny's take: I don't buy it
Manager-to-report ratios, the pastoral role, swanning around looking busy, and the market doesn't care what you enjoy

32:56 — Quick fire: Ideogram
A small, opinionated image model built for visual design — and why it points toward personal, self-hostable AI tools

38:27 — Quick fire: Midjourney's wild pivot
From sci-fi image generation to… health and wellness medical devices

43:50 — In the Weights — are you in the training data?
Tom and Jonny test themselves live. Results are humbling for one of them.

47:55 — Upcoming AI design workshop
Half-day, ticketed, late July/early August — build your own AI design tools

---

## Links:
- https://www.linkedin.com/news/story/middle-managers-in-the-firing-line-8966722/
- https://ideogram.ai
- https://www.midjourney.com/medical
- https://intheweights.com

Find us at nearfuture.works or email the pod at podcast@nearfuture.works with suggestions and ideas for content or guests.

Click here to view the episode transcript.

0:00 — Introduction & the content pipeline
Jonny and Tom kick off the episode and share their very low-tech approach to collecting topics throughout the week — a Slack thread — which immediately surfaced the biggest AI story of the moment.

1:27 — Fable 5, banned in minutes: the sovereignty wake-up call
Anthropic's most powerful model yet was blocked for all non-US citizens within four days of launch. What does it mean when critical infrastructure can be switched off by a foreign government overnight?

4:42 — Europe 2031: a cautionary tale
Jonny breaks down the 18,000-word narrative podcast europe2031.ai — a near-future story of Europe's failure to build sovereign AI capability, its dependence on open models that eventually disappear, and the grim choice between American and Chinese technology.

9:53 — What does this actually mean for you?
Tom pushes for practical takeaways. If AI vanished tomorrow it'd probably be fine — but the longer dependency grows, the harder it becomes to function without it. From job market effects to information control, the risks are subtle but compounding.

13:21 — Practical steps: diversify your AI diet
Rather than avoiding AI, Jonny argues for being intentional about where you get it from. Use open-weight models, try Mistral, explore OpenRouter — don't build critical dependency on US big tech whose incentives may not align with yours.

15:36 — The acceleration problem and European incentives
Tom reflects on how the structural incentives driving US and Chinese AI development will naturally widen the gap with Europe — even without any bad intentions from the US side.

17:18 — London rising: DeepMind, Cursor, and the talent story
Cursor just announced a 200-person London office. Tom traces how Demis Hassabis's insistence on keeping a London team when Google acquired DeepMind seeded a generation of AI talent — now paying dividends for the whole city.

19:14 — Nuance: AI isn't just bad or just good
Jonny argues for holding the tension — concentration of power is genuinely worrying, but the potential benefits to humanity (drug discovery, scientific breakthroughs) are also real. The podcast's purpose is navigation, not evangelism.

22:51 — Building a business in the AI era
The guys get into the meta-question: what is the right business model for an AI consultancy? Training vs. capability delivery, day rates vs. outcome-based pricing, what to give away free vs. what to keep — and why neither of them really has the answer yet.

35:24 — Tom's open-source project: Kazam
Tom reveals what he's been building — a lightweight open-source design framework for creating portable, lo-fi design micro-tools as single HTML files. No npm, no server, no dependencies. Inspired by a workshop where 60 designers built 60 tools in 3 hours.

Links
europe2031.ai
Jacob Heftmann
grillitype.com
Our 1 page HTML tool example: https://dither-shape-p.ooda.run/

Find us at nearfuture.works

Chapters

Building a Design Tool in 15 Minutes at Product Unleashed
Tom and Jonny recap the workshop they ran at Product Unleashed, where they built retro design tools (Etch A Sketch, MS Paint) live in front of an audience — twice. They discuss how the challenge evolved, why they pivoted from serious tools to fun ones, and how they got the final result live on the internet using a QR code.

Why Designers Building Their Own Tools Matters
Jonny shares the narrative from the workshop's opening slides — a Gutenberg-moment argument for why designers are now able to build hyper-personalized tools for the first time, referencing Adobe, Figma, and the growing landscape of AI-native design tools.

Historical Examples: Eames, Lennon & the Apple Calculator
Tom walks through three historical examples of makers building tools to unlock their own creativity: Chris Espinoza's Calculator Construction Set for Steve Jobs, Ken Townsend inventing automatic double tracking for John Lennon, and Charles and Ray Eames creating the Kazam machine to bend plywood into new furniture forms.

The Value of Craft in the Age of AI
Jonny reflects on Karri Saaranen (founder of Linear) and his view that design is the thinking — and there's no shortcut to it. The conversation explores whether something made quickly can ever be truly iconic, and why craft still has real, felt value even as AI capabilities grow.

AI as Amplifier or Counterbalance: Knowing Your Working Style
Tom introduces the idea that AI shouldn't just universally amplify — for some people (perfectionists who stall), it should act as a counterbalance, nudging them to ship. For others (fast movers who skip deliberation), the opposite. Both share personal examples of how they've tuned their own AI workflows around their personality types.

Ooda: Instantly Sharing Internal Prototypes
Jonny shares an update on Ooda, his tool that lets teams publish prototypes, tools, and documents to a private, secure link in seconds — just by asking Claude. He explains the pivot from "run your dev server in the cloud" to "don't let your team's work disappear," and why the response from a handful of people last week convinced him the problem is real.

Building a Calculator with a 5-Year-Old
Tom shares a heartwarming weekend story of building a calculator with his five-year-old son Waltie — complete with a car button, a horse button, and a Lego-themed interface — and reflects on what it means to grow up in an era where building software is as natural as playing. Jonny draws parallels to how touchscreens changed children's expectations of interaction.

• (00:00) - Episode #3:
• (00:04) - Building a Design Tool in 15 Minutes at Product Unleashed
• (07:45) - Why Designers Building Their Own Tools Matters
• (12:13) - Historical Examples: Eames, Lennon & the Apple Calculator
• (19:57) - The Value of Craft in the Age of AI
• (22:45) - AI as Amplifier or Counterbalance: Knowing Your Working Style
• (34:32) - Ooda: Instantly Sharing Internal Prototypes
• (43:59) - Building a Calculator with a 5-Year-Old

Find us at nearfuture.works

Click here to view the episode transcript.

• 0:04 — Mostly Working: AI in Design at Monzo — Tom recaps the event with James Storer, prototyping at Monzo, and why the team focuses on learning speed over shipping speed
• 6:11 — Three Use Cases for AI in Design Workflows — Jonny breaks down the emerging buckets: shipping production code, faster prototyping, and building internal tools
• 13:12 — Is Figma Losing Its Grip? — Jonny's spicy take on Figma overreaching, seat cancellation stats from a 90-person design org, and the Adobe comparison
• 18:17 — Figma's Growth vs. The Designer Exodus — Two things can be true: enterprise growth numbers and a slow professional exodus
• 21:29 — The Smartphone Camera Analogy — Did smartphones kill professional photography? And will AI do the same to design?
• 25:52 — Upcoming Events: Berlin, Shoreditch & Brighton — Jonny's "Tools That Shape Us" talk at Hatch Berlin (Sep 18), the live build session at Product Unleashed Shoreditch, and the Brighton workshop
• 35:37 — What We're Building — Jonny's prototype hosting tool, a Notion CRM, and Tom's AI-powered family holiday journal idea

Find us at nearfuture.works

Click here to view the episode transcript.

Welcome to the Near Future Podcast, where Tom and Jonny explore what they're building, learning, and experimenting with as AI reshapes how we work.

In this episode:

Agentic Murder Mystery
Jonny walks through his experiment building a six-agent murder mystery game using NanoClaw. He created a cast of AI suspects — each with their own secrets, personalities, and email inboxes — and ran a live session with a group of exited startup founders. The conversation covers what makes agents distinct from standard LLMs (triggers, tools, memory, collaboration), and the surprisingly playful emergent behaviour that unfolded when participants tried to interrogate and social-engineer the AI suspects.

The Near Future Logo Instrument
Tom shares his experiment turning the Near Future logo into an interactive, playable design tool — built in Claude Code — that generates ambient generative music as the logo animates. The conversation explores what it means to treat a brand asset as a toy, and how ephemeral software experiments fuel deeper craft practice.

Craft, Design, and What Lasts
Jonny reflects on receiving his long-awaited Poem Clock — a piece of hardware by Matt Webb that displays AI-generated rhyming couplets telling the time — as a lens for thinking about what makes design feel timeless. The duo discuss Dieter Rams' principles, the difference between craft as polish vs. craft as practice, and what that means for designers navigating the AI era.

Hiring Designers in the Age of AI
Tom shares observations from a recent head of design hiring process — how AI is adding noise on both sides, what strong portfolios look like right now, and the balance between deep foundations and pushing the frontier.

Design Engineering & the Future of the Design Team
A candid discussion on whether every designer needs to become a design engineer, the role of different archetypes within a team, and how to close the gap between Figma and code without putting everyone in an IDE.

Tools We're Excited About

• Tom: Flora AI (visual asset generation with reference images), Notion's new agent capabilities
• Jonny: UK-built sovereign design tools — Dawn Labs, Superhands, and Dessn

• (00:00) - Pilot episode
• (00:04) - Introduction
• (01:07) - Agentic Murder Mystery
• (15:17) - The Near Future Logo Instrument
• (22:27) - Craft, Design & What Lasts
• (29:03) - Hiring Designers in the Age of AI
• (33:30) - Design Engineering & the Future of Design Teams
• (45:56) - Tools We're Excited About

Find us at nearfuture.works