Who's got the ball on carbon removal?

Summary

Nan Ransohoff makes the case that durable carbon removal must be treated as infrastructure for a climate transition: emissions cuts do most of the work, but removals still have to scale now and eventually reach government-backed demand. She describes Frontier as Stripe’s advanced market commitment, a mechanism borrowed from vaccines that guarantees future demand only if suppliers deliver, helping de-risk early technologies and pull forward the best ideas. The conversation emphasizes underexplored pathways like surficial mineralization, where grinding reactive rock and exposing it to air and water can permanently mineralize CO2, potentially at very large scale if the market can tolerate longer time horizons. Ransohoff also argues that AI could both increase near-term emissions and make future carbon removal more affordable by driving growth, while the next wave of AI-linked philanthropy will require much more organizational capacity, legible problem definitions, and stronger talent pipelines. The episode broadens into “orphan problems” and “general managers,” with Ransohoff arguing that public goods need clear owners, founder-like operators, and outcome-oriented institutions.

Notes

• Carbon removal is framed as a wartime-scale mobilization problem, including talent pipeline creation and a "PM to GM" transition for public goods work.
• The IPCC 2018 report served as the field’s wake-up call; Ransohoff describes carbon removal as suddenly feeling necessary at massive scale.
• Durable carbon removal is treated as storing CO2 for roughly 1,000 to tens of thousands of years.
• The mitigation split is presented as roughly 90% emissions reduction and 10% removal, with both needed immediately.
• Frontier is described as Stripe’s advanced market commitment for carbon removal, with nearly $700 million contracted.
• Frontier’s stated portfolio target is building toward 1.4 million tons of annual removal capacity.
• Ransohoff estimates carbon removal will ultimately require hundreds of billions of dollars per year, far beyond current voluntary demand.
• The market is still constrained by the small size and time sensitivity of voluntary buyers.
• Surficial mineralization is presented as a promising rock-based pathway: grind reactive rock, expose it to air and water, and let it permanently absorb CO2.
• Ransohoff argues the pathway could theoretically scale to a teraton for less than $80 per ton, but today’s market structure under-explores it.
• Frontier’s demand signal is tech-agnostic and only pays on delivery, making it a market-shaping instrument rather than a simple prize.
• The AMC mechanism is borrowed from vaccines, with the pneumococcal vaccine cited as the first precedent.
• Stripe’s carbon removal effort started as a 2019 experiment to buy $1 million at any available price.
• Corporate philanthropy in the U.S. is large, about $50 billion a year, but often too conservative for this kind of market creation.
• The episode repeatedly returns to the bystander effect: everyone assumes someone else owns the carbon removal problem.
• Ransohoff’s selection criteria for orphan problems are world benefit, tractability, and personal commitment.
• She argues the best operators for public-goods problems look like Silicon Valley founder types: execution-heavy, strategic, and persistent.
• Existing institutions often have leaders, but the deeper issue is that many are scoped around slices of a problem rather than outcomes.
• Respiratory infections are identified as a neglected, tractable orphan problem with major health, productivity, and pandemic costs.
• A Stripe-hosted symposium with about 40 scientists and policymakers was used to test whether the respiratory infection problem was technically solvable.
• The next philanthropy wave is estimated at roughly $37 billion to $100 billion per year from AI-linked wealth creation and pledges.
• She wants a Silicon Valley-style philanthropic stack: funders, capital allocators, and many more philanthropic startups.
• The implied operating model is not a handful of donors but thousands of organizations and hundreds of thousands of workers.
• Ransohoff argues the sector needs more legible problems, higher-status on-ramps, and better compensation to attract talent.
• She sees the next 5, 10, and 15 years as decisive for AI transition risk and the definition of human flourishing.