Search for a command to run...

Timestamps are as accurate as they can be but may be slightly off. We encourage you to listen to the full context.
This episode features Anton Osyka, CEO of Lovable, discussing the explosive growth of his AI-powered no-code platform from 4 million to over 100 million ARR since his last appearance in February. (00:00) The conversation covers Lovable's mission to democratize software development, making it possible for non-technical users to build production-ready applications through conversational AI. Anton also shares his vision of building Europe's first trillion-dollar company from Stockholm, Sweden, and discusses the broader implications of AI replacing traditional coding practices.
Host of This Week in Startups and founder of multiple successful companies including Launch and Inside.com. He's an active angel investor with investments in companies like Uber, Robinhood, and Thumbtack, and operates The Syndicate, an angel investing community with over 11,000 members.
Co-host of This Week in Startups and former senior editor at TechCrunch. Alex brings extensive experience in startup journalism and analysis, having covered the tech industry for years before joining the This Week in Startups team.
CEO and founder of Lovable, an AI-powered no-code platform that has grown from 4 million to over 100 million ARR in under a year. Previously worked as a founding engineer at Sauna Labs (a recent unicorn) and served as CTO at a company that raised $20 million from top VCs before founding Lovable.
Anton predicts that AI-generated code will be twice as secure as code written by expert developers within the next 18 months. (23:03) This represents a fundamental shift in how we think about software security, where human error becomes the weak link rather than the strength. The context behind this prediction stems from Lovable's security-first approach to AI code generation, where they've implemented multiple scanning layers to ensure code quality and security. This takeaway suggests that organizations should start preparing for a future where AI becomes the primary method for writing secure code, particularly for standard applications like banking software.
Young professionals can differentiate themselves by proactively building solutions using AI tools like Lovable, even without formal approval. (37:57) Jason emphasizes how employees who show initiative by creating automated solutions or applications can dramatically advance their careers. This mirrors the historical "spec work" movement in design, but now applies to software development. The key is demonstrating value before asking for permission - building something that solves a real problem and then presenting it as a fait accompli. This approach is particularly powerful for young workers struggling to break into competitive tech markets.
Stockholm's success in creating unicorns stems from a culture of long-term thinking and engineers who focus on building quality products rather than chasing trends. (51:18) Anton explains that Swedish developers tend to stay committed to projects for longer periods, creating deeper expertise and more robust products. This cultural difference, combined with a strong technical education system and early adoption of AI research, has created a unique ecosystem. The diaspora effect from successful companies like Spotify, Klarna, and Skype has created multiple generations of experienced entrepreneurs who reinvest in the local ecosystem.
Lovable's Discord community of over 106,000 members with 8,700+ online simultaneously demonstrates the power of user-driven education and support. (43:07) Anton credits community members teaching each other advanced platform features as crucial to user success. The platform runs approximately 300 grassroots events annually, though Anton admits this is an understaffed opportunity. This takeaway highlights how successful B2B platforms need to go beyond traditional customer support to create spaces where users can share knowledge, best practices, and advanced techniques with each other.
By creating an opinionated technology stack rather than trying to support every possible configuration, Lovable can deliver more reliable results for AI-generated applications. (25:06) This approach contrasts with more generic AI coding tools like Cursor or Claude, which must work across many different architectures. The constraint paradox applies here - by limiting choices, the platform can optimize deeply for specific use cases and provide better outcomes. This suggests that successful AI tools will need to make strong architectural decisions rather than trying to be everything to everyone.