跳转至

📡 每日精华简报 — 2026-04-25

自动生成 09:30 共 3 篇


1. Alex Karnal - The Trillion-Dollar Health Revolution - [Invest Like the Best, EP.467]

来源: 🎤 Podcast (transcribed) · 87,871 chars

🎯 TL;DR: Proactive health stack investment is a trillion-dollar revolution driven by existing drugs and AI.

3 个最值钱的洞察:

  • GLP-1 drugs prove consumers are ready to pay for proactive health, unlocking a massive market beyond diabetes/obesity; Novo Nordisk and Eli Lilly are key beneficiaries.
  • PCSK9 inhibitors offer ~50% LDL reduction with minimal side effects, representing one of the best risk-reward interventions; underappreciated relative to GLP-1s.
  • Most published biomedical research is irreproducible, creating a data quality crisis; AI-first companies that generate reliable data have a huge advantage.

💰 投资视角: Beneficiaries: GLP-1 leaders (NVO, LLY), PCSK9 players (AMGN, REGN), AI drug discovery platforms (RXRX, ISPC). Damaged: supplement companies, academic-reliant biotechs. The shift to proactive health increases long-term conviction in life sciences ETFs.

📂 原文 · podcasts/invest_like_the_best/alex-karnal-the-trillion-dollar-health-revolution-invest-like-the-best-ep467.md


2. Terence Tao – Kepler, Newton, and the true nature of mathematical discovery

来源: 🎧 Dwarkesh Podcast · 74,897 chars

🎯 TL;DR: 科学发现依赖长期判断与启发式,AI目前是补充而非替代。

3 个最值钱的洞察:

  • Kepler的柏拉图立体模型虽错但推动了发现,验证闭环可长达数十年,当前AI缺乏这种长期判断力。
  • AI擅长模式匹配与生成假设,但难以理解理论深度,更多是加速而非颠覆科学发现。
  • 数学研究中人类+AI混合体将主导,因AI无法替代直觉、美学判断和长期领域积累。

💰 投资视角: 受益于AI辅助科研的标的(如微软、DeepMind、Anthropic);传统手工科研领域面临转型,AI工具提供商(如Labellbox)长期增长。

📂 原文 · podcast_blogs/dwarkesh/2026-04-25_terence-tao-kepler-newton-and-the-true-nature-of-mathematical-discovery.md


3. Quantum 101

来源: 🎧 ChinaTalk · 65,615 chars

🎯 TL;DR: 量子计算是下一个关键技术,竞争窗口窄,需立即行动。

3 个最值钱的洞察:

  • 量子计算能解决经典计算无法处理的分子模拟问题(如青霉素),将对药物发现和材料科学产生革命性影响。
  • 量子竞争赢家不单是造出最强机器,更在于制定标准、保障供应链、加密防护和产业扩散。
  • 量子计算从理论转向工程,硬件和纠错取得突破,但瓶颈转向计算机科学(编译、架构、系统集成)。

💰 投资视角: 受益标的:量子计算硬件公司(如IonQ、Rigetti)、量子软件/加密相关。受损标的:依赖经典加密的传统科技公司。对中国量子领先的担忧增加对美国本土企业的信心。

📂 原文 · podcast_blogs/chinatalk/2026-04-25_quantum-101.md