Startup City: Specials Magazine

Lars Tvede, an engineer and economist with a long career in tech and finance, has written 18 books and founded 13 companies. Throughout his work, he made a key observation: financial information was readily available and well-structured, accessible at the flip of a switch on Bloomberg terminals. However, there was no such user-friendly solution for mapping and predicting scientific and technological advancements. Navigating this crucial information always proved to be a constant hassle and mess. So, in 2019, he founded Supertrends to fix this problem. “At first, I thought we could do it in two years, but it took five”, he says with a laugh. However, it is now finished and rolling out. Supertrends: Mapping Human Innovation At the heart of Supertrends lies a comprehensive timeline of human innovation. It boasts 16,000 key innovations spanning a staggering 3.3 million years. “Creating this involved 75,000 AI prompts and queries, combined with extensive verification,” explains Lars Tvede. “But Supertrends isn’t just about the past. It predicts the future too, with 4,000 predictions currently mapped on the timeline. Each prediction is meticulously explained and placed in its proper context. And the system keeps evolving, adding an average of three new predictions every day.” To ensure accuracy and credibility, Supertrends collaborates with a panel of 160 experts – leading entrepreneurs, scientists, and industry association managers. These experts review the AI-generated outputs, refining formulations or predicted breakthrough dates. They also provide justifications for any contextual changes they make. Supertrends even taps into their insights for podcasts and blog content through scheduled interviews. Behind the Scenes: A Symphony of Automation Mapping the current scientific and technological landscape involves a sophisticated blend of methods. Every day, an algorithm meticulously scans approximately 5,000 leading science and technology media sources in 30 languages, searching for 3,000 specific keywords. A natural language processing (NLP) model then summarizes these daily findings into around 300 unique feature stories. Other AI models eliminate redundancies within these stories and automatically check if any of them align with the predicted innovations. If a predicted innovation finds real-world confirmation, the system automatically reclassifies it from a prediction to an actual event. As Tvede aptly states, “Millions of small, automated routines run the system every day.”

What if you could make a payment—even with insufficient funds—without relying on a credit card or racking up interest? That’s the promise of SmartPay, a company redefining digital finance in Japan. Today, when Buy Now, Pay Later (BNPL) options are often tied to hidden fees, complex terms, or high-interest credit cards, Smartpay has carved out a new path— one that’s clean, transparent and radically simple. At the core of Smartpay’s offering is a fully digital, one-click financial ecosystem—the first of its kind in Japan. It allows users to split payments into three equal installments over two months, all without any interest or added fees. This isn’t just about deferring payments; it’s about enhancing affordability and giving consumers greater control over their spending without the usual risks that come with traditional credit. What makes Smartpay truly different is its frictionless experience. By linking directly to your bank account, the system ensures automatic, on-time deductions—eliminating the need to juggle due dates or manually make payments. There’s no dependency on credit cards, no compounding debt and no fine print. Just a seamless way to manage cash flow with confidence and clarity.