Wild Hearts – Détails, épisodes et analyse

If you're here for commentary on the Pope, Trump, or the geopolitics of frontier AI - this isn't that episode. If you're here for the unfiltered view from the people actually building Claude - stay.

This one is for the builders, the tinkerers, and the curious. Two of the people behind Claude - not here for governments, the press, or the Vatican. Instead, here to zero in on what they're seeing, and how you can get more from AI, however you're using it.

Katelyn Lesse runs engineering for Anthropic's Claude Developer Platform. Angela Jiang runs product. They're the people closest to what builders are actually doing with the technology and are exactly the kind of spark people they'll tell you every team needs.

What they're seeing: teams that transform overnight because one person in them is genuinely obsessed. Founders who move with the model instead of against it. A shift from "content is king" to "context is king" that most people haven't caught up to yet. And a clean slate advantage - available to anyone willing to look at an old problem as if it's never been solved before.

Earlier this year, Eucalyptus sold for $1.6 billion - one of the biggest exits in Australian startup history. Charlie Gearside co-founded it.

A year on, he's spending his time on YouTube and on Build Australia, the nonpartisan movement he launched last month to make space for a more ambitious version of this country. He talks publicly about property investing, the universities, and a culture that calls earnest people try-hards. None of it is comfortable, and none of it is what most people do after an exit.

He's also the first to admit that none of this pays off quickly. The work is to shift what Australians think is acceptable to want. A project that’s measured in decades, not quarters.

Kate sits down with Charlie to talk about untangling your identity from a company you built. The brutal question he had to ask himself before leaving Eucalyptus. Why he won't go into politics. Why he thinks the most risky thing right now might be doing nothing. And the one topic he's still too nervous to make a video about.

There’s a graveyard of robotics companies—billions torched on beautiful demos we’ve all seen before, but never felt. This episode explains why the economics, the software, and the demand curve have finally flipped—and how Alloy plans to fuel the winners.

Joe Harris returns to Wild Hearts—but this time as a founder. An engineer by training (ML for telecoms), operator by practice (Eucalyptus growth & product), and obsessive systems thinker, Joe unpacks why robotics is finally crossing from hype to inevitability. We trace the structural shifts powering the moment—collapsing hardware costs, foundation-model intelligence, and urgent customer pull—and the hard lessons from failed vertical farming plays that recalibrated what reliable automation actually demands. Joe introduces Alloy, a horizontal data and observability platform for robotics teams: find the 1% of mission data that matters, surface edge cases, track reliability toward “four-, five-, six-nines,” and shorten the loop from failure → fix → redeploy. If you’re building, buying, or betting on robots, this is the market map and playbook for the next decade.

• The three real drivers: cost curves, capability (VLM/VLA), and customer pull
• Reliability as the business model: why 99% isn’t enough—and how teams get to 4–6 nines
• Data, not demos: robots emit GB/min; how to isolate the 1% that changes outcomes
• Horizontal vs. vertical: what failed in indoor/vertical farming and why
• Alloy’s wedge: multimodal search (images, time series, logs), “scenarios,” alerts, and instant mission summaries to accelerate deployment and reduce unit costs
• Team & culture: hiring for speed, humility, and learning in a field moving weekly

00:00 First operator-to-founder return: Joe’s path (engineer → Atlassian → Eucalyptus → Alloy)

02:00 Maker roots: coding tutorials at 12, early internet leverage

03:30 Many small businesses → the “one-thing, 10–20 years” decision

08:30 Why now for robotics: cost curves + reusable rockets as mindset shift

10:45 Vertical farming post-mortems: unit economics, reliability, scale errors

13:40 Reliability is everything: from 99% to 99.999% in the physical world

15:45 The data firehose: GB/min, multimodal chaos, and missing tooling

18:40 Operator-to-robot ratio as the core unit economic lever

21:10 Selling into robotics: design partners, security, and data heterogeneity

23:15 Common data primitives (perception, time series, logs) + ROS-driven formats

24:30 Why LLMs aren’t enough: context-window limits & multimodal encoding

27:00 Alloy’s product: natural-language search, similarity, “scenarios,” real-time alerts

28:50 Instant mission summaries vs. days of manual analysis

29:30 Edge AI tailwinds: Jetson class hardware, cheaper sensors (LiDAR/IMUs)

30:30 VLAs explained: from perception → plan → act (and why smoothness matters)

32:10 The pace of change: weekly breakthroughs, staying on the frontier

33:40 Distribution & adoption: enterprise first; consumer follows reliability

35:40 Safety and necessity: underwater, heavy industry, logistics

37:15 Autonomy acceptance: the “first Waymo ride” unlock

43:00 Ideal customers: high throughput, real deployments, cloud telemetry

44:50 ICP discovery playbook: questions that qualify real readiness

45:50 Team design: missionary talent, humility > hubris, learn-fast culture

46:40 Macro lens: robotics as a deflationary lever & company formation boom

48:00 Jobs & leverage: from decoding info → higher-order coordination

50:05 The Alloy analogy: the coal-shoveler that keeps the engine running

You don't have to be the founder to build the future.

When Andrea Quinn made the leap from fashion merchandising to tech, she didn't start a company. She joined one. Today, she's VP of Go-To-Market Operations at Halter, New Zealand's newest unicorn, which just raised $155 million at a $1.55 billion valuation.

Not every path into building the future looks like a founder origin story. Some of the most crucial work happens when you join the right company at the right moment and help turn ambition into execution. Andrea's doing exactly that - scaling the GTM motion as Halter accelerates across Australia and the United States.In this episode, Blackbird Partner Sam Wong sits down with Andrea to explore how operators translate skills across industries and build the engines that power billion-dollar companies. From her Commercial Equation framework to practical AI applications in sales, Andrea breaks down what it actually takes to scale a startup from the inside.

This episode is for: founders building GTM, operators inheriting messy funnels, and anyone wondering if they need to start a company to build the future.

Because the answer is no. Sometimes the most valuable thing you can do is join the rocket ship and help build the engine.

When storytelling meets startup energy, magic happens.

In this week’s episode, Xavier Collins, co-founder of Wonder, joins Mason to explore how technology is tearing down the old gates of Hollywood, and what happens when anyone, anywhere, can tell stories that move the world.

Backed by Blackbird and LocalGlobe, Wonder is building an AI-native creative studio reimagining how films are made, who gets to make them, and what “production” even means. Xavier shares how AI can help the 90% of scripts that never get made finally see the light of day - from resurrecting forgotten footage to helping bold new voices get their first break.

We dive into instinct versus analytics, courage versus consensus, and the scrappy startup mindset redefining creative industries. It’s a story about belief, innovation, and the people daring to create what others think impossible.

This episode is for anyone who’s ever had a story they’ve wanted to tell, a dream they’ve wanted to build, or an idea they’ve been told was too crazy to work.

Because when content becomes infinite, the only thing that matters is the quality of the story - and your story might just be next.

When Michelle Battersby launched Sunroom, she set out to change the game for women creators, building a platform where they could earn freely, safely, and on their own terms. Five years, three funding rounds and one pandemic later, she did just that. Thousands of creators made life-changing income, and Sunroom was acquired by Fanfix.

From the emotional weight of leadership to the surprising financial realities of building something from scratch, Michelle shares the unfiltered truths of the founder journey - the highs, the hard parts, and the freedom that comes with letting go. Maddy Guest, from Blackbird’s investment team and host of the finance podcast So Invested, joins Michelle to unpack what those lessons teach us about resilience, risk, and redefining success.

This is a story about ambition and endurance — and the lessons that only reveal themselves when you decide to climb.

When Lucy Liu co-founded Airwallex in 2015, she was flying around the world opening bank accounts in person and carrying bags of security tokens. Global businesses are digital. But finance was stuck in the past.

For three years, Airwallex burned money building invisible infrastructure no one believed in yet. Her co-founder drew a “really ugly unicorn” on a whiteboard predicting ten-times growth when they had zero revenue. Everyone laughed - but beneath the laughter was a serious undertone that they were onto something big. Something that would be game changing. So they kept building.

That bet on infrastructure became one of the fastest-growing fintechs in the world, now moving $200 billion annually and adding $100 million in recurring revenue every quarter.

In this episode, Lucy shares why building two products simultaneously defied conventional startup wisdom, how hiring for intellectual curiosity beats credentials, and what it means to scale from zero to 1,800 people without losing speed. She also reflects on the power of ambitious predictions, staying simple at massive scale, and why resilience matters more than perfection.

Procreate co-founder James Cuda has spent more than a decade obsessing over one thing: the brushstroke. From hacking the iPad 1 to run at 60fps, to turning a side project into the world’s leading creative app, James has built Procreate on a radical philosophy: simplicity, permanence, and creative freedom above all else.

In this episode of Wild Hearts, James joins Mason to share why the company never took VC money, how “flow state” shapes everything from product design to team culture, and what it really takes to scale without losing soul. They also dive deep into generative AI, ethical data, and why Procreate’s biggest unfair advantage may simply be staying small and Tasmanian.

James also reflects on the tension between addition and reduction, the power of jam sessions, and why listening to the “little voice” is the artist’s greatest superpower.

Time Stamps

00:00 – Intro

02:05 – Why brushstrokes were the starting point

05:10 – The art of subtraction: keeping flow while adding features

07:50 – Permanence as a product philosophy

09:36 – From “an amazing piece of shit” to a world-class creative tool

12:11 – How Procreate’s archetype grew from amateurs to architects

15:01 – Listening to users without losing the soul

17:31 – Scaling creativity and protecting flow inside the team

19:51 – Jam sessions, “holy shit” moments, and making ideas real

23:31 – James’ strong stance on generative AI and ethical data

34:51 – Authenticity over slogans: building trust with artists

37:21 – Bringing artists together, online and offline

39:06 – Staying independent: why Procreate never took VC

44:01 – Simplicity vs. optionality in future workflows

46:39 – The advice James gives every artist: listen to the little voice

48:26 – Outro

What if you could take the most mysterious force in physics—and make it useful?

In our final  episode of this season of Wild Hearts, we sit down with Pete Shadbolt, co-founder of PsiQuantum, a company racing to build the world’s first utility-scale quantum computer. But this isn’t a conversation about quantum theory. It’s about execution. Engineering. Scaling. Building something that moves humanity forward - not in decades, but now.

Pete shares why 300 or 3,000 qubits won’t cut it, and why a million is the magic number. We explore the technical marvels (and madness) involved in the team’s journey: superconducting detectors millimetres from red-hot heaters, lasers brighter than a trillion photons, and a cryostat that throws out the chandelier model altogether.

But most of all, this is a story of ambition. Of leaving behind prestigious academic careers, raising a billion dollars, and assembling a team of physicists, welders, aerospace engineers, and cryo-specialists to take one shot at building something historic.

In this conversation, we cover:

🚀 Why PsiQuantum is chasing 1 million qubits—not 300, not 3,000🏗️ What it takes to move quantum computing from theory to hardware—with welders, chip designers, and aerospace engineers 

📉 Why academia can be a trap—and how PsiQuantum built an anti-academic company culture 

🌐 The real-world applications of quantum computing: from designing drugs to revolutionising materials science 

👩‍🔬 How team DNA, not just tech, shapes PsiQuantum’s ability to scale and execute 

⚙️ Why quantum computing isn’t a mass adoption tool - and why that’s perfectly okay

🔥 How engineering targets that once caused mutiny are now being hit daily

This episode concludes our fifth season of Wild Hearts. Over the past 40 weeks, it’s been our honour to chat to the founders and operators shaping the world we live in. If you’ve enjoyed the conversations, we would be grateful if you could like, subscribe, and share our program with other wild hearts. 

Wild Hearts will take a short break, and will return to all streaming platforms later this year. 

From everyone at the Wild Hearts team, thank you! 

What if planning dinner wasn’t a chore—but something you looked forward to? In this episode, Wild Hearts guest host, Silk Kadala - investor at Blackbird - chats with Anna Guerrero, founder of Clove, a beautifully designed cooking app that’s reimagining how we cook at home.

You might know Anna from her nine years scaling the creator marketplace at Canva—but it was a stint as a pasta chef in the Dolomites that ultimately set her on the path to launching Clove.

Whether you’re interested in the role of AI in reducing decision fatigue, why brands are betting big on recipe creators as the next wave of culinary entrepreneurs or just stood in front of the fridge thinking “what’s for dinner?”—this episode is for you.

🔍 In this conversation, we cover:

🍳 The invisible mental load of everyday cooking—and how Clove is removing it with Smart Planner

📲 Why Clove’s approach to AI is more whisper than shout—and why that matters for creativity

📚 Building for creators: how Clove is giving food bloggers, TikTok cooks and chefs a new way to publish and earn

🎯 From pitch decks to real traction: Anna’s high-stakes decision to pause Clove’s creator program and set a new quality bar

🚀 The leap from Canva exec to culinary school student—and what working in a Michelin-starred restaurant taught Anna about product

🧠 Low ego, high initiative: what Clove looks for in early team members and building a culture of adaptability

🧭 What it means to follow the dots—why you don’t need to have it all figured out to move forward

🍽️ The long-term ambition: turning Clove into the global go-to for “what’s for dinner?”—with a billion recipes cooked through the platform

From Canva to Clove, Anna Guerrero shows what it looks like to reinvent yourself, back a bold vision, and build something that truly changes how we live and cook.