@sama "governments decide to offtake a lot of computing power and get to decide how to use it, and it may make sense to provide lower cost of capital to do so"
offtakes = guaranteed demand = lower cost of capital = infrastructure financed + built = secure supply
AI can only automate a task if the result can be checked easily
The bottleneck for many valuable problems is not solving the task, but verifying whether the output is correct--that is the next wave of opportunity in AI
on the "asymmetry of verification" @_jasonwei
On Jan 25, 2025, BlackRock released a report that marks a watershed moment for natural capital.
The world's largest asset manager has officially recognized natural capital as an investment factor, signaling a systemic shift in market valuations.
https://t.co/zqTlVf5Z62
7/
Classic models (public goods, Hotelling, Rosen/Roback,...) have limited spatial-environmental links.
Space is often "here" or “anywhere else”; cities have limited geography.
Example: is Clean Air too lenient? NYC/Boston no closer than NYC/LA
Pic from https://t.co/SiO2vhz1at
Imagine a VPPA… but for low-carbon materials. Companies not offtaking the cement can buy the environmental attributes associated with the lower carbon product. We’re seeing the emergence of a “book and claim” system for low carbon materials. https://t.co/voUy7LO1Cq
The primary energy fallacy: Critics of the energy transition often point to primary energy to demonstrate that the transition is going to be impossible.
But most consumed primary energy is lost after conversion.
The problem to solve is a lot smaller than primary energy suggests.
refrigerants have 10,000X the global warming potential of CO2 per ton. yeah.
curious about entrepreneurs working on the refrigerant crisis?
check out my friend jamie’s article!
Who'd start their first company in a country they’d never visited to solve an unsexy climate problem that nobody ever thinks about? @louispotok, that's who.
Inspired by his audacity, I flew to SE Asia to visit and write his story. Read it now (link in🧵)
Daylight Ecosystems placed 2nd at @OpenAI x @lowercarbon x @CrusoeAI hackathon this weekend! We map the ecosystem of climate experts and decision makers to accelerate decarbonization 🌍
So excited to participate in the @OpenAI x @lowercarbon x @CrusoeAI hackathon and win 2nd place!🏆
For this AI x Climate hackathon my team and I built a tool to give energy project developers instant insight, through a knowledge graph approach, into the permitting review process
@nickvanosdol huge. and here I was getting hyped about DR-EAF.. although DR-EAF using hydrogen as a reductant in iron ore + renewables abundant places like Australia makes sense