All Sessions：

針對生成式 AI 的發展歷程以及其對台灣中小企業的影響進行了深入探討。首先，回顧了生成式 AI 的三個發展階段：從 2020 至 2023 年的爆發期，進入 2024 年的理性調整期，並展望 2025 年的成熟應用期。這一過程中，AI 技術從炒作到逐漸實用化，尤其是多模態模型和 AI 代理人（AI Agents）等新技術的崛起。演講的核心部分聚焦於如何通過 AI 工作流程和 AI 代理人來提升企業效率。例如，介紹了設備借用自動化系統，通過數位化流程來解決傳統設備借用管理中的痛點，從而節省大量的人力與時間，並將這些資源重新投入到更具創新性的任務中。此外，會議中展示了如何使用 Google NotebookLM 進行 RAG（檢索增強生成）技術的應用，從而進一步提高工作效率。演講的另一重點是未來 AI 技術的發展趨勢，尤其是從 AI 輔助到 AI 代理的過渡。這種變化將從最初的任務輔助演變為更深層次的協作夥伴關係，促進人類與 AI 的協同工作。 最後，AI 的導入不再是一個選擇題，而是企業生存的關鍵。為了幫助中小企業快速適應 AI 技術，他提出了「三步走策略」（Crawl-Walk-Run）：從策略規劃、快速驗證到全面擴展，逐步推進 AI 工具的導入。演講總結指出，數據是 AI 應用的燃料，文化的建立與團隊學習成長是成功的關鍵。企業應該從最小的痛點著手，通過低成本的 AI 工具進行實踐，並在實踐中不斷優化，最終實現人機協作，創造更大的商業價值。

Over eons, synaptic connectivity endowed animals’ carbon-based brains with high-dimensional processing for survival and decision-making. We now approximate the perceive–understand–decide pipeline with ANNs and multimodal models, while modern silicon handles large, high-dimensional data within explicit energy and latency budgets. Building truly human-like autonomous systems still faces critical bottlenecks, demanding disruptive architectures to maintain consistent performance in dynamic, unpredictable environments on a few-tenths-of-a-watt power budget. Many autonomous systems still center on a compute-intensive, centralized “System 2,” paired with a low-latency sequence-generation engine (the so-called “System 1”). While fine for labs or simulations, this design makes balanced trade-offs among reliability, size, mass, and energy efficiency nearly impossible; the absence of proprioception impedes long-horizon consistency, robustness, and adaptability. Shift from centralized compute-first to layered, bottom-up autonomy centered on distributed proprioception: low-latency, low-power perception–action loops (the real System 1) handle immediate responses; a predictive, adaptive brain (System 2) governs strategy. Evolutionary distillation and related methods optimize each subsystem’s energy efficiency, mass, volume, reliability, and safety. Leveraging strengths in semiconductor manufacturing, advanced packaging, and embedded-system integration, Taiwan is well positioned to drive and define the low-power, high-volume sensing and actuation chiplet markets, building up a complete supply chain of “neural reflex systems” (low-latency, real-time reaction subsystems) for the autonomous systems and robotics industry. This will not only reduce dependence on ultra-high-compute “brains” but also strengthen Taiwan’s position in the global ecosystem ahead of large-scale commercialization of autonomous robots and intelligent systems, and enhance its strategic influence in future end-to-end system competition. 經過漫長的演化，動物的大腦靠突觸把神經細胞連起來，慢慢長出「高維度資訊處理」的能力，才能在複雜環境中存活與決策。對應到今天，我們用人工神經網路（ANNs）與多模態模型，在功能層面近似「感知→理解→決策」的流程；同時，現代矽基硬體也能在明確的能耗與延遲預算內，處理大量高維度資料。 但要做出真正「類人」的自主系統，還卡著關鍵瓶頸：要在動態、難預測的場景裡長時間保持穩定表現，而且功耗只有二十多瓦，必須採用更顛覆的新架構。 目前許多自主系統仍以計算密集、集中式的「System 2」為核心，再配一個低延遲的序列生成引擎（所謂「System 1」）。這在實驗室或模擬環境或許可行，但一到實務現場，就很難同時兼顧可靠性、尺寸、重量與能效；而且欠缺本體感知（proprioception），使得長期一致性、穩健度與適應力都受限。 解法是從「算力集中優先」轉到「由下而上、以分散式本體感知為核心」的分層自主架構：由低延遲、低功耗的感知—動作融合迴路（真正的 System 1）負責即時反應；上層由更會預測、能調適的「大腦」子系統（System 2）負責策略治理。配合演化式蒸餾等方法，各子系統可分別在能效、質量、體積、可靠性與安全性上做到最佳化。 台灣的優勢在這裡特別關鍵：我們結合半導體製造、先進封裝與嵌入式整合，有機會主導低功耗、高出貨量的感測與致動 chiplet（小晶片）市場，建立面向自主系統與機器人產業的「神經反射系統」（低延遲、即時反應子系統）完整供應鏈。這不只減少對超高算力「大腦」的依賴，也能在大規模商用化之前，強化台灣在全球生態系的地位，並提升未來端到端系統競爭中的戰略影響力。

Manufacturing faces urgent global challenges – rising costs, supply chain disruptions, labor shortages, and mounting pressure to decarbonize. In Taiwan, the manufacturing sector accounts for over 50% of national CO₂ emissions, making it a critical priority in the journey toward Net Zero. This presentation explores how Siemens’ Digital Twin and Industrial AI technologies are empowering manufacturers to achieve more with less – boosting efficiency, reducing emissions, and building greater resilience. Through real-world reference cases, we’ll show how these technologies are already transforming factories – from virtual design to intelligent operations – driving a new era of sustainable, smart industry in Taiwan and beyond.

Based on observations from AI transformation projects in industries like retail and manufacturing over the past two years, the core challenges of building AI applications have evolved into the "Agentic" era. This presentation explores the journey from developing basic RAG (Retrieval-Augmented Generation) applications to engineering complex Multi-Agent Systems. We will break down the key challenges faced, the practical engineering methods that work, and the modern architecture required to support these advanced systems.