AI Will Change Humanity Faster Than You Think | Bohan Lou

Leonard Pickard
Loading transcript... Click for full transcript
About this episode Bohan Lu, partner at Chemistry.vc, discusses venture capital as a study of human judgment under extreme uncert… AI summary

Bohan Lu, partner at Chemistry.vc, discusses venture capital as a study of human judgment under extreme uncertainty, emphasizing that early-stage investing requires betting on founder agency and perseverance over static business models. He analyzes the competitive landscape between US and China, noting China's dominance in open-source AI models and rapid infrastructure build-out, while the US retains advantages in talent, capital markets, and compute resources. The conversation also explores the transformative potential of reproductive technologies, longevity science, and the societal implications of AGI.

Key takeaways 6
  • Venture Capital Risk Profile: The industry follows a power-law distribution where approximately 50% of investments may go to zero, 30% break even or go sideways, and only 20% perform well, with one 'unicorn' potentially paying for the entire fund. Top-tier firms like Sequoia target a 40-50% write-off rate to ensure sufficient variance for outlier returns.
  • Founder Traits Over Intelligence: The best founders are not necessarily the rawest geniuses (like theoretical physicists) but possess 'aggression' (a hunger to win), commerciality (focus on economic value), and the ability to avoid intellectual drift by focusing narrowly on their business goals.
  • China's AI Infrastructure Advantage: China is building a self-sufficient semiconductor industry and supercomputing infrastructure to bypass US export controls. This centralized mobilization allows them to produce highly efficient, sparse AI models that are currently the most popular open-source models used by Silicon Valley startups.
  • US Structural Advantages: Despite China's progress, the US maintains three key advantages: a continuous 'brain drain' of global talent, deeper and more risk-tolerant capital markets, and superior access to compute resources necessary for large-scale AI training.
  • Reproductive Technology Frontier: We are entering an era of 'reproductive technologies' including embryo selection, germline editing, and in vitro gametogenesis. This could reduce the generational cycle from years to weeks and allow for selective breeding for traits like intelligence and disease resistance, raising significant ethical questions.
  • Longevity Escape Velocity: Ray Kurzweil predicts 'longevity escape velocity' by 2032, meaning life expectancy will increase faster than aging occurs. Demis Hassabis (DeepMind) predicts solving all human maladies within 10-20 years. Chemistry.vc invests in companies like Isomorphic Labs and Fathom Therapeutics working on computational drug discovery and undruggable targets.
Notable quotes 5 AI-generated: wording and quote attribution may be wrong. Use the play link to verify.
  • “I always think of venture capital as the study of human judgment under extreme uncertainty.”
    Bohan Lu defining the core nature of his job as a VC.
  • “The best founders are not always the smartest people by raw intelligence... What they also have is aggression... They have to want to win. They have to want to compete and they have to have a hunger in them to build a venture scale business.”
    ▶ 12:03 Describing the specific personality traits required for startup success beyond raw IQ.
  • “China's outcome now... lifting the most humans out of abject poverty in human history in the shortest amount of time was very much not the modal based case... I feel like he [Deng Xiaoping] could be a great founder.”
    ▶ 8:06 Comparing historical leadership traits to startup founder archetypes.
  • “The whole open AI stack of the world is powered by China... Chinese open source models are the most popular AI models used by Silicon Valley startups even more than the anthropic open AI models.”
    ▶ 10:55 Highlighting the irony that US restrictions have led to Silicon Valley relying on Chinese open-source AI tools.
  • “Life is too short to live any other way. So young people give it the big swing. Go for the stars.”
    ▶ 56:06 Final advice to young people about taking risks and exercising agency.

Chapters & Sections (28)

0:02 Venture Capital Judgment and Uncertainty chapter 1
3:02 Venture Capital Judgment Under Uncertainty
5:29 Venture Capital Risk and China Comparison chapter 1
7:38 China's Technological Advancement and US Structural Advantages
10:19 Chinese AI Dominance and Founder Traits chapter 1
12:20 Founder Traits and Investor Due Diligence
14:51 Investing in Founders and AI Underestimation chapter 3
16:40 Intellectual Stimulation and Investor Curiosity
18:12 Respectful Rejection and Founder Relationships
19:39 AI Accelerationism and Superintelligence Engagement
22:09 US and China AI Advantages chapter 2
23:55 US AI Advantages: Talent, Capital, Compute
25:18 China's Centralized AI Mobilization vs US Public Opposition
27:05 AI Energy Demands and Reproductive Tech chapter 2
29:18 Space-Based Solar Energy and Data Centers
30:40 Reproductive Technologies and Genetic Selection
32:53 Genetic Selection and Longevity Escape Velocity chapter 1
35:31 Longevity Escape Velocity Predictions
37:31 AGI Arrival and Existential Questions chapter 1
39:40 AI Safety, Science Fiction, and AGI Beliefs
41:54 AI Impact and Human-Machine Symbiosis chapter 1
45:15 Mental Usurpation and Human-AI Symbiosis
46:26 AI Safety, Bioweapons, and Human Regrets chapter 5
48:54 AI Guardrails and Bioweapon Prevention
50:03 Reflecting on Regrets and Overthinking
51:53 AI in Drug Discovery and Biology
53:51 AI Drug Discovery and Future Advice
55:04 Overcoming Stress and Embracing Agency

Transcript

Loading transcript...