Fireside Chat, Interview
The Future of Nuclear Energy -- Vinod Khosla
- The speaker argues that forcing non-economic climate solutions is ineffective at scale because energy is a massive-volume problem where subsidies must decline as market penetration grows; without becoming economically competitive, technologies cannot be broadly adopted.
- Early-stage subsidies are justified for the first 1–5% of the market to aid development, but by the time a technology reaches 5% market share, it should no longer require subsidies to remain viable.
- Political and economic realities in developing nations like India and China make it unfeasible to trade immediate economic access (e.g., clean water, energy) for long-term climate benefits, meaning solutions must be cheaper than fossil alternatives to be adopted.
- The speaker asserts that incremental solutions are insufficient and that the industry requires "large breakthrough technologies" (100x scalability) to truly displace existing energy sources.
- The speaker's investment bets for future energy lie in fusion and deep geothermal (400–500°C), which offer the necessary scalability that incremental site improvements cannot provide.
- Venture capital investments made over 10 years ago in companies like QuantumScape, Verdeji, Lanzatec, and Impossible Foods have returned 5–10x multiples, indicating success in patient capital for hard tech climate solutions.
- Current portfolio companies are producing new cement that is 70% less carbon than traditional Portland cement and is already cheaper, with operational plants scheduled for the current year.
- Additional breakthrough areas of focus include new plant proteins, new fertilizer production techniques, and radically different approaches to steel and cement production.
- Fusion technology became viable due to high-temperature superconductors, which enabled the creation of 20 Tesla magnets.
- The scaling relationship of the reactor is proportional to 1 over Tesla to the power of 4; quadrupling magnet power reduces reactor size by a factor of 50 (250 times smaller than previous iterations).
- Commonwealth Fusion Systems built a 20 Tesla magnet and demonstrated it in 3–3.5 years, a pace 10x faster than the 25–30 years required for large international projects using smaller magnets.
- The speaker projects that demonstrating positive net energy from fusion could occur by 2025–2026, also representing a 10x acceleration over historical timelines.
- Commonwealth Fusion instigated a broader field shift, attracting over a dozen credible fusion startups and top talent, increasing the probability that one or two will succeed.
- Widespread adoption of fusion, geothermal, and new extraction methods is expected to reduce dependence on oil and minerals, thereby decreasing international tensions and strategic conflicts over resources.
- Scaling fusion requires a "modular" deployment strategy where existing coal or gas boilers are replaced every 5–10 years with fusion boilers rather than building 5,000 new power plants; this would require constructing 5,000 boilers, a logistical feat comparable to WWII Liberty ship production.