Johan Ståhle

Most pre-seed rounds in the Nordics follow a familiar pattern: About $1.5M, usually led by a dedicated local pre-seed fund. These rounds are important for the ecosystem. They provide founders legitimacy, support, and a starting point. But pre-seed investing isn’t only about solid foundations. It’s also about extremes. The best founders often don’t fit neatly into the middle of the curve. Some raise <$1M syndicates: fast, scrappy, "headless" (h/t Sarah Drinkwater) rounds that bring multiple great people (angels, micro funds, Solo GPs etc) around the table, providing early validation before a breakout seed. Others raise $3M+ internationally led rounds, skipping the incremental step and going global from day zero. In 2024, I thought I wanted to build a pre-seed fund. But then I realised most pre-seed funds live in this middle while the best companies were being built at the edges. That’s where I now focus: either bringing a syndicate around a diamond in the rough, or earning my place as the rough in the diamond! If you draw out the typical size of pre-seed rounds, it looks like a bell curve. Pre-seed, as it’s usually practiced, is about the middle. I invest in the tails. That’s why I no longer think of my fund as a “pre-seed fund” but rather a "first-cheque fund" operating at the edges of the earliest days. 

228

14 kommentarer

Employee + AI Collaboration = A delicate balancing act like a seesaw perched on a pea. There's a crazy tension in the heart of the new wave of AI / Human teamwork Sometimes AI can lighten the load and spark more creativity and productivity Other times it can crush productivity and leave teams feeling uninspired and frustrated --- THIS IS THE CHALLENGE --- Give employees smart AI partners, and on paper everything improves: Routine tasks handled automatically, freeing up mental space. Workload drops, with less cognitive and psychological strain. More energy for creative leaps, innovation, and going the extra mile. But according as this paper illustrates beautifully the tension is real Offloading work to AI does unlock proactive behaviors like process improvement and ideation but mainly when employees feel confident using the tools. Without enough AI literacy, the benefits flatten or get weird: workers with lower AI skills may stay proactive, but their behavior is more about “staying in the loop” than reaching new heights. --- WHY IT’S TRICKIER THAN IT SEEMS --- On the surface, it looks like a win/win: Employees do fewer repetitive tasks. Proactivity and innovation metrics rise. Dig deeper, though, and you find nuance: Employees with high AI literacy get the biggest relief, able to “give away” tedious work and reinvest their effort in higher-order challenges. Those with low AI literacy tend to react more uniformly; sometimes ramping up initiative, but often from a place of uncertainty over job relevance or a lack of trust in automation. And when the workload drops but employees don’t feel equipped or involved? The boost in proactivity can stall; or even backfire, leading to disengagement or stress. --- HOW COMPANIES CAN GET IT RIGHT --- Target Workload, But Preserve Ownership Don’t aim for “full automation.” Use AI to take the grunt work, but keep employees engaged in the tasks that fuel their purpose and sense of accomplishment. People need to feel their work is meaningful to take real initiative. Customize AI Onboarding and Training Invest in tailored support: Offer advanced options for AI-fluent employees. Provide structured, hands-on training for those less confident. One-size-fits-all won’t work: the impact of AI is as much about skills and comfort as the tech itself. Design for Active Collaboration Make AI a collaborator, not a replacement. Build workflows where people decide, review, and create—rather than just approving what the AI does on autopilot. Encourage proactive “co-piloting” instead of passive oversight. --- THE BOTTOM LINE --- There’s no guaranteed shortcut from AI adoption to a culture of proactive, creative employees. The most successful teams will be those that focus not just on what gets done, but how empowered people feel while doing it. Human / AI collaboration doesn’t mean handing over the keys it means letting employees drive, with intelligent tools riding shotgun

113

9 kommentarer

Swedish AI startup Lovable, a platform that lets users build apps and websites using plain language, has secured $200M (nearly €172M) at $1.8B valuation in a Series A round led by Accel. Read more here: https://lnkd.in/dFWEN36b 20VC, byFounders, Creandum, Hummingbird Ventures, Visionaries Club #SiliconCanals #News #Funding

55

Sweden loves AI: First Klarna, then Spotify, and now Sweden’s Prime Minister admits he uses ChatGPT “quite often” at work 😳 Not for speeches. Not for documents. But for a second opinion. He asks things like: → “What have others done in this situation?” → “Should we think the opposite?” No secret files uploaded. Just raw perspective. This is fascinating, especially since AI often provides faster and sharper answers than human advisers. More importantly, studies show AI-assisted policy can be 15% more accurate. And this isn’t just Sweden. Leaders worldwide are quietly leaning on AI, from city mayors drafting budgets to ministers brainstorming policy. I guess the question now is not if politicians will use AI. It’s whether they’ll admit it.

391

84 kommentarer

Absolutely insane week in startup world: - Cognition acquires Windsurf, beating out Google and OpenAI in a high-profile bidding war - Lovable becomes Europe’s newest unicorn: just eight months since launch, the Swedish AI "vibe‑coding" startup raised a $200M Series A led by Accel, at a $1.8B valuation - Perplexity AI is now valued at $18B after a fresh funding round, up from $14 billion just two months ago - Thinking Machines Lab, founded by ex‑OpenAI CTO Mira Murati, raised a record-breaking $2B seed round at a $12B valuation - Several top‑tier VCs launched new funds this week, including Brian Singerman, raising over $500 million for GPx, a fund backing emerging VC managers - Substack raised $100M, reaching a $1.1B valuation, with backing from a16z & others - Astronomer CEO Andy Byron was caught in an affair on camera with his HR chief at a Coldplay concert - OpenAI rolled out the new ChatGPT Agent Stay up to date with all of it in the weekly EverythingStartups newsletter. Link in my bio.

171

14 kommentarer

This tiny team is behind the world’s fastest-growing startup. They hit $17M in 90 days. Now they’ve launched something in Europe that’ll disrupt every business that ignores it. Here’s the wild story of Lovable AI: In 2023, two Swedish engineers asked: What if anyone could build real software by just typing in plain English? No code. No dev team. No friction. Just: I need a task manager with user auth and a dashboard. Boom—working app. They called it Lovable AI. 90 days later? 30,000 paying customers. By April 2025: $50M ARR — growing at 2,816% YoY. So what exactly does it do? It’s not another no-code toy. It’s an AI-native app builder: → You describe the app → Lovable builds it → Fully functional, fully editable → Frontend, backend, logic done → Deploy in seconds Under the hood: → React + Vite frontend → Supabase backend → LLMs from OpenAI, Gemini, Anthropic → GitHub integration for prod deploy Production-grade speed. The builders: Anton Osika – Co-founded Stockholm AI, award-winning AI systems. Fabian Hedin – Built Stephen Hawking’s comms interface, worked with ex-SpaceX engineers. VCs noticed: €14.3M pre-Series A from Creandum (Spotify), Thomas Wolf (Hugging Face), Charlie Songhurst (ex-Microsoft), Adam D’Angelo (Quora, OpenAI board). But... cracks show. → Custom logic? Unreliable. → Debugging? Painful. → Edge cases? Flaky. Every action burns credits—when something breaks, you’re paying to fix it. Then came the security bombshell. In April 2025, researchers gave it a 1.8/10 security rating. They auto-deployed phishing sites on Lovable’s subdomains—credential harvesting, malware dashboards live. An indie hacker’s dream. An enterprise’s nightmare. The takeaway: Lovable is not for fintech or sensitive data. But for: → Solo founders → Indie hackers → Startups racing to MVP ...it’s nitro. Traction: → 500,000+ users → 25,000+ apps daily → 4.7★ on Product Hunt It’s not perfect. But for speed-to-launch? It’s magic. If you care about long-term security, look elsewhere. Final lesson for founders: Don’t hide behind your tech. People buy from people. Your personal brand isn’t vanity— It’s leverage. It’s trust. It’s survival. Founders: We’ll build your personal/company brand on X (and beyond) without you lifting a finger. We've already helped 40+ Founders get 1.2 billion combined views. Interested in how we can do this for you? Book your free discovery call here: https://lnkd.in/eUKpRuRW

39

13 kommentarer

Stockholm-based Lovable has hit over $100 million in annualized revenue in just eight months by using AI to enable millions of non-coders to instantly turn their ideas into websites, apps and online side hustles. (Photos: Sebastian Nevols for Forbes) https://lnkd.in/eEqUC4q3

529

14 kommentarer

We should all watch Varjo’s ironically named documentary, VR is Dead. It looks beyond the hype of immersive technology and reveals what’s really happening in the industrial space. I constantly hear misconceptions about AR, VR, and MR - both positive and negative - and I can tell you that 99% of it is hype. But there’s much more going on here than most people realize. I had the opportunity to try Varjo’s USD $10,000 VR headset at a conference in California last year, and I honestly couldn’t put the experience into words. But this kind of technology will become more widely accessible in time - and time will tell. Here is the documentary: https://lnkd.in/g6SP3G38

303

30 kommentarer

I have a love/hate relationship with the fertility vertical in femtech. On one hand, I'm happy to see advancements and innovation for people hoping to grow their families. But on the other hand.... where's the corresponding investment for maternal and infant health? Fertility solutions have received the most investment across femtech, with over $10 billion in venture and private equity investment in the past decade. We need the same attention and investment focused on maternal health where the maternal mortality rate is still 18.6 deaths per 100,000 live births in the US, and in some states the rate is as high as 50 maternal deaths per 100,000. We know what we need to do to fix this! In my recent Fast Company article, I share the three ways technology can help drive down maternal mortality and give women and babies the best chance for long, healthy lives: 📉 Real time pregnancy monitoring and screening 🛑 Preterm labor detection and prevention 👶 Postpartum care A huge thank you to all of the organizations investing in this tech: Laerdal Million Lives Fund, Medicines360, Catalyst by Wellstar. And special shout out to all of the startups in Femovate who are driving change forward: Kim Smith, FACHE of HealthEvolve, Meghan Doyle of Partum Health, Kelsey Mayo of Armor Medical Inc, Roni Cantor-Balan of Ciconia Medical, Andrea Cubitt of Dionysus Health, Sheila Pande of Tandem, the whole team at BB Imaging, and Steve Porter and Susan Mann at riskLD Inc. Link to the article in the comments

110

23 kommentarer

This 40-person Swedish startup hit $100M ARR, and just raised $200M at $1.8B valuation - only 8 months after launch. Here's what's behind their success: The startup - Lovable. It’s an AI platform that lets you build software using only a chat interface. They started in September 2023 and now have 2.3 million users and 180,000 paying subscribers. In my opinion, these the 3 factors made this explosive growth possible: #𝟭 𝗣𝗲𝗿𝗳𝗲𝗰𝘁 𝘁𝗶𝗺𝗶𝗻𝗴 They spotted what ChatGPT meant for software development early, and acted on it immediately. The market was in their favor and they seized the opportunity – whether by chance or by strategy. #𝟮 𝗣𝗿𝗼𝗱𝘂𝗰𝘁-𝗹𝗲𝗱 𝗴𝗿𝗼𝘄𝘁𝗵 Their free plan adds a "built with Lovable" badge to projects, creating viral, organic distribution every time a user shares their app. #𝟯 𝗘𝘅𝗽𝗮𝗻𝗱𝗶𝗻𝗴 𝘁𝗵𝗲 𝗺𝗮𝗿𝗸𝗲𝘁 𝟭𝟬𝟬𝘅 Less than 1% of the global population can code. Lovable lets the other 99% build software by simply talking to AI. This expanded their addressable market from dev teams to literally everyone with an idea. I’ve seen this shift firsthand at Waterglass.ai. Before: we built client interfaces in Google Sheets because it was the fastest option. Now: with Lovable, v0, or Cursor, we can create polished applications faster than setting up a spreadsheet and connect it with our custom-build backend. Lovable’s success only proves something I couldn’t be more happy about: The future belongs to solopreneurs and small, focused teams solving real problems – with AI doing the heavy lifting. What’s your take?

35

11 kommentarer

Lovable became the fastest company to ever reach $100M ARR And it wasn’t built the traditional way -  that’s exactly the point. When I first heard about Lovable, a Stockholm-based #startup claiming to turn natural language into working apps, I was intrigued. I’ve spent enough time in early-stage venture and working with technical founders to know: → bold claims are easy, execution is everything. But then I watched it happen in real time. 8 months after launching, Lovable crossed $100M in ARR. Faster than OpenAI, Wiz or any software company I’ve ever seen. Just 45 people, millions of users, and a product that spoke directly to a deep, immediate need: → make #software  creation as intuitive as speaking your mind. And they delivered. From where I sit - working closely with early-stage founders every day - I see 3 big reasons why Lovable got there so fast: 1. They nailed the user mindset- not just the use case. Lovable didn’t build for #developers.They built for the millions of people who wish they could build like developers. Designers, founders, operators - people with ideas but no way to ship. The product met them exactly where they were, and let them dream bigger. 2. They made #AI practical, not just impressive. This wasn’t a showcase of what AI could do - it was a tool that did the job. Lovable’s "vibe-coding" interface, and later, their autonomous agent, saved time, reduced bugs, and actually shipped production-level work. AI as leverage, not spectacle. 3. They focused, relentlessly. What’s wild is how small the team stayed - just 45 people. That takes discipline. It also takes clarity of vision. From what I’ve seen, the startups that grow fastest aren’t the ones doing the most. They’re the ones doing the least, but doing it incredibly well. Do you think this kind of speed is the new normal or still the exception? Drop your thoughts in the comments! ~~~ Enjoy this? ♻️  Repost it to your network and subscribe on Substack https://lnkd.in/gEKRbt6p for more news on #AI, #Investing and  #Innovation 🧠

52

8 kommentarer

Denmark is about to make history with groundbreaking proposed legislation that would give every citizen copyright ownership of their own face, voice, and likeness. This isn't just policy—it's a fundamental reimagining of identity protection in the AI era. What Denmark is proposing: → Automatic ownership rights to your digital reproduction → Legal power to demand takedown of non-consensual deepfakes → Severe fines for platforms that don't comply → 50-year post-mortem protection Why this matters: Current systems are broken. Most AI platforms just block real people entirely because of legal uncertainty. Denmark is showing how to solve this properly—by giving individuals control and creating frameworks for ethical AI collaboration. At Official AI, we've always operated on this principle: consent, credit, and compensation should be the cornerstone of AI content creation. Denmark's proposal would give this approach legal teeth. The contrast is striking: Denmark treats NIL rights as urgent personal liberty, while the U.S. relies on a patchwork of state laws with no comprehensive federal framework. This isn't anti-technology—it's pro-human agency in a technological world. When individuals have clear ownership rights and platforms provide tools to exercise them safely, we get the best of both worlds: incredible AI-powered creativity with sustainable consent and compensation. Denmark is creating a template for the world. Read the full analysis: https://lnkd.in/gqhszr2p #AI #CopyrightLaw #NILRights #Denmark #TechPolicy #DigitalRights #OfficialAI

41

5 kommentarer

Vibe-coding won’t create 1-person billion-dollar companies. But it will create thousands of 1-person million-dollar ones. I’m in Sweden, where I spent the weekend with my younger brother (he works in construction), vibe-coding utility apps for his industry. Using Replit and off-the-shelf AI tools, we prototyped two MVPs while having a few beers. A year ago, the fidelity of these prototypes would have taken me at least six months of weekends and nights—it now took us a Friday evening. AI maximalists often say, “One person will build the next billion-dollar company.” The problem with that statement is the copies. If everyone has the world’s best AI in their pocket, copying brilliant ideas won’t be hard—or slow. It’s the luxury industry’s copycat problem, but at the speed of SaaS. This is already happening in the AI content creator space. Instead, I think we’ll see a different trend—something closer to a Flappy Bird effect: A wave of one-person, million-dollar apps. Built fast and solving niche problems. Modest revenue (but significant to an individual) And the second they rise above the radar: copied to death. And that's fine - it will then be time to move to something new. Micro-utilities for niche fields. Tools for your cousin the plumber. Your aunt the dentist. Your neighbor the beekeeper. If you have ideas and know people with real jobs, this is your moment. Fewer excuses. More experiments.

36

8 kommentarer

Many first-time founders get one thing dangerously wrong when scaling - and it can cost them their whole company: Culture. Not because they don’t care about it, but because they think it’s something that just happens when you hire good people and keep the energy high. And that works… Until it doesn’t. What holds things together at five people - shared Slack jokes, late-night product pushes, unspoken understanding - starts to fall apart at 50. By the time you hit 500+, the cracks widen into gaps that slow you down, burn people out, and chip away at the thing that once made your team great. Culture is how decisions get made when you’re not in the room. It’s not your perks. It’s not your values’ slide. It’s the invisible system shaping how people communicate, take ownership, handle conflict, and move under pressure. Here’s what the best founders understand: ✅ Culture is the most powerful flywheel for scale and success.  ✅ Culture is infrastructure. ✅ Culture is how incentives are structured. ✅ Culture is how transparency, trust, honesty, and other virtues are applied each day: first by the founders, then by each new hire individually. Culture is the sum of individual, day-to-day decisions.  ✅ It’s not what you say: it’s what gets rewarded, what gets tolerated, and what gets ignored. ✅ It starts with how leaders show up - and scales through how frontline managers lead when no one’s watching. Founders don’t scale by being in every room. They scale by embedding clarity - through culture. That means defining it early. Reinforcing it constantly. Hiring people who don’t just “fit,” but expand and strengthen what you’re building. And yes - treating roles like Head of People or Head of Talent as critical infrastructure hires, not back-office support. When culture is intentional, onboarding gets faster, silos shrink, and execution becomes sharper. You don’t need to micromanage, because expectations are built into how the company runs.

95

18 kommentarer

“AI won’t take your job.” They said... Meanwhile, in March 2024, Klarna announced their AI assistant was doing the work of 700 full-time employees. That’s not a futuristic scenario. It's already happening now. "𝗧𝗵𝗲 𝗮𝗱𝗼𝗽𝘁𝗶𝗼𝗻 𝗼𝗳 𝗔𝗜 𝗮𝘁 𝗞𝗹𝗮𝗿𝗻𝗮 𝗵𝗮𝘀 𝘀𝗶𝗴𝗻𝗶𝗳𝗶𝗰𝗮𝗻𝘁𝗹𝘆 𝗶𝗺𝗽𝗮𝗰𝘁𝗲𝗱 𝗶𝘁𝘀 𝗵𝗶𝗿𝗶𝗻𝗴 𝘀𝘁𝗿𝗮𝘁𝗲𝗴𝘆. 𝗧𝗵𝗲 𝗰𝗼𝗺𝗽𝗮𝗻𝘆 𝗲𝘅𝗽𝗲𝗿𝗶𝗲𝗻𝗰𝗲𝗱 𝗮 𝟮𝟱% 𝗿𝗲𝗱𝘂𝗰𝘁𝗶𝗼𝗻 𝗶𝗻 𝗵𝗲𝗮𝗱𝗰𝗼𝘂𝗻𝘁 𝗯𝘆 𝘁𝗵𝗲 𝗲𝗻𝗱 𝗼𝗳 𝟮𝟬𝟮𝟯, 𝗙𝗼𝗿𝘁𝘂𝗻𝗲 𝗿𝗲𝗽𝗼𝗿𝘁𝗲𝗱." So while we were told to “work alongside AI”… #AI already clocked in, learned the job, and started working. Let’s be honest: Yes, AI is a great tool. Yes, it can boost productivity. Yes, there are exciting ways to use it. But let’s not pretend this isn’t about economics. If a tool can do the work of hundreds for cheaper, faster, and without benefits do you think companies won’t use it? We don’t need empty optimism. We need: → New skills  → New policies  → And new questions How will we fit into the future of work? Maybe it's time we stop repeating the nice slogans and start asking harder questions. Because if we’re not asking those questions now. We may not like the answers later. Do you still think AI won't take your job? #FutureOfWork #AI #LinkedInVoices #KlarnaAI #DigitalWorkforce #ThinkForYourself

86

77 kommentarer

Quote from Andrew Ng that nails what we’ve felt at Skarbe in 2025: “If you can build a prototype in a day but wait a week for product feedback, that’s your real bottleneck” This is exactly what happened to us in 2024/2025. With gen AI, we can ship features insanely fast. But what slowed us down, and still does, is deciding what’s worth building at all. We’re now at almost 300 customers. Every week, I get a flood of raw, unfiltered feedback. It’s constant. And honestly, that’s what’s shaping the product way more than anything we imagined at the whiteboard stage. My biggest lesson: -> Product owners are now the bottleneck, not engineers. -> You can ship features in hours. But figuring out what actually solves a real problem (and what to say “no” to) takes deep customer empathy and fast decision-making. Most days, that’s on me + Alex (co-foudner and CTO) I’ve spent 6 years as a PM and product leader in B2B saas unicorn, thought I “got it.” But nothing humbles you like hundreds of real users giving feedback, often pulling you in different directions. It’s not about collecting votes or shipping requests. It’s about feeling the customer’s pain and quickly turning that signal into a better product. What works for us now: -> Building tight feedback loops: weekly check-in calls, chat groups, emails, real voices, not just data. -> Giving the product owner real power to say no, focus, and act fast. Waiting for consensus = dead end. -> Shaping the roadmap in real time, not quarterly, not monthly, but actually IN REAL TIME! Bottom line: -> AI makes building fast. The real work is listening, deciding, and shaping. Every. Day. -> Product ownership isn’t just a job, it’s the constraint for every AI startup right now. Any folks out there feeling this shift as you scale?

26

3 kommentarer