Sunghee Yun (He/Him)

Co-Founder & CTO @ Erudio Bio, Inc., Co-Founder & Leader & Chair of Silicon Valley AI Nexus (AI Nexus), Advisor @ Korean American Semiconductor Professional Alliance (KASPA), CGO / Global Managing Partner @ LULUMEDIC, KFAS-Salzburg Global Leadership Initiative Fellow @ Salzburg Global Seminar, AI-Korean Medicine Integration Initiative Task Force Member @ The Association of Korean Medicine, Director of AI Semiconductor @ K-BioX, Visiting Professor @ Sogang University, Advisory Professor @ DGIST, Global Advisory Board Member @ Innovative Future Brain-Inspired Intelligence System Semiconductor of Sogang University, and Network Expert @ Gerson Lehrman Group

Abstract

Artificial Intelligence (AI) has moved from research curiosity to civilizational force in a little over a decade, and this talk sets out to separate genuine capability from surrounding hype for an audience of ambitious young scholars. It traces the arc from the Deep Learning (DL) revolution of the early 2010s—AlexNet, AlphaGo, the Transformer—through the large language model (LLM) and generative AI (genAI) explosion of the 2020s, to the agentic AI systems now going mainstream. Along the way it grounds the story in hard indicators: adoption curves outpacing every prior platform shift, staggering private funding, and hyperscaler capital expenditure crossing the trillion-dollar threshold in 2026. The aim is not to dazzle but to give students an honest map of where we actually are in the AI lifecycle—still, arguably, at the dawn.

The talk then examines what makes modern AI work and where its real frontiers lie. It explains why LLMs became such powerful representation learners—arguing that natural language, refined by evolution and used to hand down human knowledge for millennia, offers an exceptionally rich structure for machines to learn from—and how multimodal systems extend this by fusing text, image, audio, and video. From there it unpacks agentic AI – the shift from systems that answer questions to systems that pursue goals through a perceive–plan–act–observe loop, built on tool use, memory, and long-horizon planning. It surveys the fast-moving agentic tooling landscape across open-source and the major labs, and makes a key architectural point—that the language model is the reasoning engine, not the whole system, and that value is migrating from raw models toward system design.

Against this backdrop, the talk delivers its central message to students – the most durable advantage lies at the intersection of deep domain expertise and AI fluency. It distinguishes two equally valuable but distinct paths—the AI power user, a domain expert who wields AI as a force-multiplying tool, and the AI expert, the researcher or engineer who builds the systems themselves—and argues that the highest-impact breakthroughs come when the two collaborate, with the domain expert leading. Concrete cases anchor the claim, from AlphaFold’s Nobel-winning marriage of structural biology and deep learning to cancer-diagnostic AI where the physician still makes the final call. The old assumption that tech companies would build AI and disrupt everyone else has given way to a new reality in which AI is a commoditized tool and domain judgment is where lasting value accrues.

Crucially, the talk insists that technical superpower is only as good as the character behind it. Because AI amplifies whatever intentions its wielder brings, the speaker argues that integrity, empathy, service, and a genuine moral compass matter more than ever—the things AI cannot do for you, and the choices that determine who you become. This framing reflects both the pace of change—where the frontier a student trains on today will have shifted by graduation, making the ability to continually relearn the frontier the truly durable skill—and the responsibility that accompanies rapidly expanding capability.

Drawing on the speaker’s own trajectory across semiconductor R&D, e-commerce AI, and now AI-driven biotechnology, the talk closes with grounded illustrations of these principles in practice—including Erudio Bio’s fusion of physics-based measurement with AI for drug discovery and diagnostics, and the Silicon Valley AI Nexus community bridging innovation ecosystems. The overarching invitation to the Jigwan scholars is clear: build genuine depth in a field you care about, learn to collaborate fluently with AI, stay adaptable as the frontier moves, and let a strong moral foundation guide how you use the power you accumulate.