During a live panel discussion on clean energy demand, Mike Schroepfer argues that general electricity demand growth in the U.S. is not driven solely by AI data centers.
Insight
Schroepfer: Customers buy savings, making 'climate tech' a confusing category label
“I would actually say it even slightly stronger, which is we use this word climate tech like it's an actual good definition of a market segment. When we just talked about like, I sell products to consumers that save the money. I'll sell products to consumers th…”
Prediction Not checkable as stated
Schroepfer: Well-funded tech companies will build massive power capacity globally
“Underhyped in the sense that I like Respectfully disagree with Nick, which is just look at NVIDIA's earning calls. Like people are buying those chips and powering them up. So, and I think companies are, I spent a lot of time being incredibly creative about fin…”
Prediction Open · timeframe May 2030
Schroepfer: Economically viable fusion will produce history's largest companies
“And so I think this is a very binary thing, that if we can actually hit unit economics that are competitive in fusion, It'll create the largest companies we've ever seen.”
Prediction Open · timeframe May 2030
Schroepfer: Nature-based carbon removal will hit sub-$100/ton surprisingly fast
“I think we will be shocked at how fast, low cost, particularly nature-based carbon removal systems, and when I say nature-based, I don't mean planting trees. I mean, enhanced rock weathering. I mean, ocean alkalinity enhancement. I mean, biochar. These are sys…”
Opinion
Schroepfer: Consumer demand response is overhyped due to fragmentation and friction
“I'm gonna say consumer sort of demand response is overhyped. It's just, like, so damn hard to get consumers to sign up for these things, and it's such a fragmented Market that getting to reasonable scales is hard.”
Insight
Schroepfer: AI does not solve climate tech's manufacturing scale-up bottleneck
“AI for material science, great for discovering new materials. The manufacturing scale-up is usually the hard part. So great, I've got this new MOF. I've got to make a lot of it very cheaply. The AI doesn't really help me very much with that. Like, in theory, m…”