“Climate Week should be rebranded to Get Energy Faster Week.” Tesla co-founder and Redwood Materials CEO JB Straubel said it on September 24, and it summed up both Houston and New York. AI load growth was the organizing force of Climate Week 2026. Almost every room we were in came back to the same questions: who builds the power, how fast, and who pays for it.
The mood among investors was sober but not bearish. The “tourist capital” has left, capital is concentrating in fewer themes, and the bar for a raise is higher. For investors anchored in molecules and firm power, that is a better setup than it sounds.
Climate Week confirmed what we have been telling our partners all year: we are in an unprecedented energy expansion, and speed to power now trumps everything else.
We think about it as the energy expansion stack. Applications, models and chips get the headlines and the capital. But every layer above depends on the one at the bottom: power. Power powers everything. That foundation is where ECV invests: Green Molecules® innovation that helps natural gas, and the firm resources that complement it, deliver the energy expansion faster, cleaner and more reliably.
Sophie Purdom called this the best year yet for climatetech, but said deployment has been flat for about three years. New energy and data centers have gone from roughly a third of new capital to two-thirds. John Rapaport noted data centers account for 40–50% of total debt issuance, dwarfing energy overall.
Andrew Beebe argued the climate bill pushed capital toward non-venture areas and drew in “tourist capital” riding the incentives. With the tourists gone, he said, it is time to get back into climate. LP follow-on is thin and capital is concentrating in fewer logos.
Rajesh Swaminathan said big rounds now need a full early-to-late investor stack from day one, citing Khosla’s $100M in Mariana Minerals’ $300M round. The takeaway: early-stage capital should lean non-consensus, and let later-stage investors consolidate around winners.
Where ECV fits: the largest venture firms are now making their biggest-ever energy bets, and a16z is running podcasts on natural gas. We were non-consensus on this thesis in 2021. Backed by gas utilities, we have stayed focused on the natural gas value chain ever since.
Affordability was the political backdrop all week. Exelon CEO Calvin Butler, asked whether the grid can be strengthened without raising prices, answered: “No, you can’t” (Heatmap). He added that the tech industry “lost the narrative up front.”The
One panelist framed it well: this is a local permitting problem more than a technology problem. Communities offered jobs, flat rates and clean power still say no. AI profits will have to be shared more directly with host communities.
With gas turbines backordered and interconnection queues long, developers are bringing their own molecules.
Load flexibility is the industry’s counter-argument on cost: Google, Nvidia and Emerald AI launched an AI energy management alliance during the week.
Where ECV fits: speed to power is the gap our portfolio is built for. Teragen Energy is developing a metal-supported solid oxide fuel cell that is smaller, cheaper and load-following, running on natural gas, hydrogen or ammonia for behind-the-meter power. Sapphire Technologies’ turboexpanders generate electricity from pressure drops on existing gas lines and is expanding ever faster in the data center world.
The supply chain behind the grid is stretched thin (Latitude Media).
Copper is the other constraint. DexMat’s CEO put it sharply: the AI build-out “will hit a wall on megatonnes of copper before it hits one on megawatts.” Its Galvorn conductor is made from natural gas. On critical minerals more broadly, the refrain was consistent: it is a processing problem, not a resource problem.
Where ECV fits: the energy expansion needs more copper. The Red Metal Company applies the shale and enhanced-geothermal toolkit (horizontal drilling, engineered stimulation, reservoir modeling) to recover copper in situ at brownfield mines that are already permitted and built, with no smelter. enaDyne’s cold plasma catalysis abates hard-to-treat emissions from semiconductor fabrication, another pinch point in the AI build-out.
At B Capital and Cerberus Ventures’ “Subsurface to Space” session, Star Catcher CEO Andrew Rush described a $60M program for space-to-space power beaming with government as first customer. Government is roughly half the space market.
He laid out three roles for government capital: catalytic funder, market maker (price floors that remove demand risk), and lender or loan guarantor. Nuclear fuel recycling came up as a case in point: the government already carries billions a year in spent-fuel liabilities, so it has every incentive to fund a solution.
First-of-a-kind financing was a thread across Node’s FOAK Summit and CREO’s Developer U. The focus has shifted from direct equity to matching each risk to the right provider: insurers, EPCs, banks and development capital. Higher rates make that discipline more important; CTVC noted the Fed’s first rate hike in over three years, driven partly by energy inflation.
Where ECV fits: Actual gives infrastructure funds and developers AI-powered capital planning, so financiers can commit earlier to multi-billion-dollar projects. Eclipse Energy turns depleted oil reservoirs into energy assets, producing hydrogen in situ and expanding into methanogenesis and de-souring on existing oil and gas infrastructure.
“19th century laws, 20th century infrastructure and 21st century problems” was the line of the week on water. Water is scarce, volatile and hyper-local, so solutions have to work across many markets and watersheds. The open question is where the data and the margins are.
Resilience framed S&P Global’s flagship forum, Tailwind Futures’ new taxonomy (physical AI, sensors and drones now a category of their own), and NYU Urban Future Lab’s first adaptation prize track.
Where ECV fits: Capture6 pairs carbon removal with water recovery, using electrodialysis to clean and desalinate industrial waste streams. Water is the lead value proposition.
What was quiet: hydrogen barely came up. Carbon markets focused on delivery risk and verification: Relae and Microsoft added a delivery-risk framework to their CDR criteria, and ExxonMobil won Texas approval for its $5B CCS project.
Houston ran the week before New York and felt different: less capital markets, more field tours and pitch stages. Organizers framed it as “more energy AND fewer emissions; reliability AND affordability; speed AND durability.”
The read-across: both cities talked about the same things — AI load, the grid, geothermal, water. New York was the capital conversation; Houston was the operators and corporates who will buy and deploy it.
Thank you, Activate Houston: thanks to Jeremy Pitts and the Activate team for having us for a candid session with their fellows during the week.
Power powers everything.
.svg){kind=avatar}