SV–Asia Youth Innovation Leaders Summit 2026

First on stage was Chih-Yao Wu, who opened with a "top 10 of the 100 greatest inventions" poll: the light bulb, the internet, the telephone, paper, the wheel, the airplane. World-changing inventions sound out of reach, but his whole point was the opposite: inventing is a skill you can learn.

The method is TRIZ, a system that distills a massive body of patents into 40 inventive principles. He showed the full table of 40, then a "bag of tricks" slide revealing how they hide in everyday objects. The tear notch on a snack bag is Preliminary Action, the upside-down ketchup bottle is Inversion, spiral mosquito coils and curved mascara brushes are Spheroidality, car turn signals are Periodic Action, and folding chairs are Dynamics. The little designs around us all follow principles.

One genuinely practical patent tip: when you find a patent in your way, don't detour yet. Check whether it has expired or lapsed because the annuities went unpaid; many "blocked" paths are already public domain. His society also runs a podcast sharing invention success and failure stories, and organizes stages like the Taiwan Innotech Expo and the IIIC international invention competition.

He closed with two book recommendations: "Fifty Questions on Creative Thinking for Children" and "The Inventor's Handbook". As for me, I fell straight down the TRIZ rabbit hole and turned the 40 principles into a bilingual field-guide site; see "follow-up deep dives" below.

Ray Chen strung his talk together as a series of questions, walking the audience through the turning points of his own life. The first one landed hard: can a mid-life crisis still open a second curve? On the slide were two photos from 2006: his young family, and Muhammad Yunus receiving the Nobel Peace Prize with Grameen Bank. That was the year microfinance showed him that business could answer social problems directly, and the start of his own second curve.

Question two: beyond profit, what else can business change? He traced the B Corp movement's arrival in Asia, from the book he co-translated ("B Corp: the good companies we need now") to the group photo of an Asia B Corp gathering, turning "good company" from a compliment into a movement with standards and a community.

B Current Impact Investment positions itself as "a lever for sustainability, a bridge for cross-sector collaboration": pooling capital from the National Development Fund, corporates and family offices on one side, and connecting startups, transforming incumbents and cross-border teams on the other, with investment as the fulcrum.

The most memorable slide for me was the Spectrum of Capital: from finance-only to impact-only is a continuum, not a binary. He summed it up as ABC: A, do no harm (screen out tobacco, gambling, weapons); B, be transparent and fair (do ESG well); C, offer solutions (address the SDGs head-on). Impact investing isn't charity; it's consciously choosing where your capital stands.

He closed on young people and AI. By the research he cited, the global impact investing market is set to roughly triple within a decade toward the trillion-dollar mark; younger investors care whether investments match their values, not just returns; and AI can amplify impact measurement and capital allocation, as long as its own energy footprint is part of the math. After this talk I turned my homework on impact investing and B Corps into a follow-up site; see "follow-up deep dives" below.

David Wu laid out his cross-industry career on a single "Investment Journey" slide: consumer electronics, e-commerce, F&B, mobile apps, live streaming, fintech, and finally venture capital, with the logos of companies he has invested in or mentored filling the screen. His career is that timeline, stepping into a new industry every few years.

The case he spent the most time on was TaxiGo (later Line Taxi / LineGo). Taxis are a licensed, concession-based industry: medallions are often run by families across three generations, so the real barrier is structural, not technical. On top of that, Taiwan Taxi (55688) had acquired most of the small fleets, leaving newcomers to face a highly consolidated market.

So how do you test product-market fit there? The playbook at the time was free-ride subsidies to spike demand. It worked, but it burned an enormous amount of cash. Subsidies buy you usage data; they don't buy a lasting moat. It was the most honest, felt piece of startup reality in the whole afternoon.

His closing slide was a country-by-country table of representative unicorns: OpenAI and SpaceX for the US, ByteDance for China, Coupang for Korea, Mercari for Japan, Grab for Singapore, Canva for Australia… and a blank cell next to Taiwan's flag. Why no Taiwanese unicorn yet? His answer wasn't capital or technology but education: raising people who dare to think big, take risks and pull resources together matters more than subsidizing any single industry.

Andy Lin opened with a single line: "Have you ever had that experience, where one day, everything you've built goes to zero?" The room went quiet. He spent the rest of his time answering it with his own story.

His path was not a fast one: NTU class of 2001, to the US for UC Irvine in 2004, ASE in 2006, then two decades in Silicon Valley before founding Yo-Kai Express. The slide showed where it stands today: seven countries, 300-plus AI-powered autonomous food stations, with SoftBank, IPPUDO and Pulmuone as strategic investors.

But the spine of the talk wasn't the wins. It was a timeline titled "failure is the mother of success": in 2013 his first startup, a grocery-delivery platform for Chinese communities in the US, failed; in 2018 he hit the wall of Japan's supply chain, where no one wanted to help a foreigner; in 2020 COVID wiped revenue to zero in a month; in 2021 Korea's suppliers shut their doors, and he kept ringing doorbells only to be turned away; in 2024 the boba texture wasn't right, so he recalled every bubble-tea machine and re-engineered it through generation after generation.

After every failure came one of his lines. After shutting down the delivery platform in 2013, wondering if he was simply "one of those people not built for startups", what pulled him back was "the thing you can't stop thinking about is the thing you truly want to do". The boba do-over slide carried "you can fail, but you cannot compromise with yourself". And running through the whole talk: "you may fail, but you must not stop trying".

From absolute zero to seven countries, the talk really said just one thing: you will fail; don't be discouraged. He has lived that sentence into proof.

Tina Cheng's talk was literally titled "Your life is not a straight line", and her résumé is the proof: a UCLA civil engineering degree, then finance (JP Morgan, investment banking), a nightlife venture along the way, then six startups, including Jigocity, an e-commerce platform sold to a listed company for US$65 million. Her major never decided her destination.

Her first framework was three circles: Passion × Expertise × Market Demand, with your "holy grail" at the intersection. The line that mattered most was the caveat: "You don't find it first and then start; you discover it through action."

Framework two: stackable, transferable skills. Stackable means each skill amplifies the next; financial analysis plus user insight plus digital marketing compound into something hard to replace. Transferable means negotiation, storytelling and data literacy travel across industries and roles. Her own formula, "engineering logic × financial thinking × consumer brands × e-commerce marketing", is what she called the foundation of six startups.

Careers, in her telling, are step functions, not lines. Banking trained her financial mind, engineering gave her systematic problem-solving, and every failed venture was a data point; each step stacks chips for the next, and none is wasted. She normalized failure in one breath: "If it works, keep going; if it doesn't, analyze why, then move forward."

The AI section had the line that stuck with me most: "AI is the co-pilot, not the pilot: the steering wheel stays in your hands." AI can generate a business plan in 30 seconds, but judging, verifying and setting direction remain your job. Her six most durable skills for the AI era: critical thinking, innovation, taste, cross-domain integration, communication and persuasion, and networking. Then the "unfair advantage": ask yourself why me, and why now. Your background, languages, culture, network and lived experience form a moat AI can never copy.

Her closing advice was unexpectedly practical: "Asking for help is the most underrated superpower." It's strategy, not weakness: be specific about what you need, do your homework first, give back what you learn, and follow up. And the young have a natural edge, because people genuinely want to help promising young people. The only thing you can lose is the chance you never asked for.

Closing the summit was Jonathan Lin, whose twenty years span operating companies, investment banking and private equity: CFO of AT&T Taiwan, chairman and CEO of a major beverage company, head roles at a global drink chain, investment banking at Deutsche Bank, UBS and Credit Suisse, then partner at Summitview Capital, Dawei Capital and Andra Capital. At Wharton he studied technology and biotech management, where, as he put it, his professors and classmates were mostly healthcare experts and CEOs; in a room like that, the network is part of the curriculum.

The clearest thread of the talk was his taste in deals: he likes backing companies that change an industry and unseat its leader, and working on era-defining M&A. The case list on his slides spanned three decades, from AOL and Time Warner, Comcast's acquisition of AT&T's cable business, and the Hong Kong IPOs of ICBC and Alibaba, up to the Musk ecosystem's Twitter take-private and the merger of X into xAI, plus a full page of Taiwan deals: Fairchild and System General, MediaTek and NuCORE, the E Ink acquisition that made Amazon's Kindle possible, and Taiwan High Speed Rail's pioneering offshore convertible bond. Behind every deal, a reshuffling of an industry's competitive structure.

He went deep on one current case: NVIDIA's roughly US$20 billion acquisition of Groq in December 2025. The deal used an unusual "asset purchase plus non-exclusive license" structure to lower antitrust risk; in substance it was an acqui-hire, buying the IP outright and folding the core engineering team in; strategically, Groq's LPU patches the GPU's weakness in real-time, low-latency inference. One transaction, played simultaneously on the technology, legal and talent boards: era-defining M&A, happening in real time.

As for what he does now: Andra Capital focuses on the world's top pre-IPO software and AI companies, and he mentioned Google as an investment partner. My biggest takeaway from the closing talk: at the top of the investing world, judgment and networks are themselves a form of capital. After the summit I did a round of independent homework on Andra Capital and turned it into a site of its own; see "follow-up deep dives" below.