Episode Summary
Executive Summary: Shail Khan discusses 2026 energy-transition themes: the unexpectedly fast, bipartisan backlash to data centers; limited direct EV impact from the Iran war but stronger European energy sovereignty; Bloom Energy’s resurgence in behind-the-meter power; the possibility of data-center tariffs subsidizing other customers; bullishness on solar and grid buildout; and why the grid/utilities may be underappreciated in AI-driven electricity demand.
Main Topics: Data-center backlash becomes bipartisan (Priority: 5/5): Khan says community opposition to data centers surfaced faster and more broadly than expected, cutting across party lines and showing up in pauses and moratoria in places like New York and Texas. Iran war and energy markets (Priority: 4/5): He argues the war’s direct effect on EV adoption has been limited because oil prices didn’t spike as high as feared and EV trends are being driven more by policy and regional factors than by the conflict alone. Europe’s push for energy sovereignty (Priority: 4/5): Khan highlights that the Iran war and Russia-Ukraine conflict are reinforcing Europe’s desire to control its own energy destiny, benefiting domestic and regional resources over import dependence. Bloom Energy as a behind-the-meter winner (Priority: 5/5): He frames Bloom Energy and fuel cells as an underappreciated beneficiary of AI power demand, especially where gas turbines are constrained and air permitting is a bottleneck. Utility tariffs and cross-subsidies for data centers (Priority: 5/5): Khan explores whether small utilities with large new data-center loads could create tariffs that significantly subsidize other customers’ electricity prices, potentially even by 20% to 50% in some cases. Solar growth and C&I solar challenges (Priority: 4/5): He remains highly bullish on U.S. solar growth, arguing forecasts still understate buildout, while noting that commercial and industrial solar needs both better economics and major soft-cost reductions. Grid and utility capacity in the AI era (Priority: 5/5): Khan’s hot take is that the market underestimates the grid’s eventual ability to scale; despite slow decision cycles, he expects far more grid-connected capacity to be built over the next two years.
Key Arguments: Data-center opposition is not just a progressive issue; it is local, bipartisan, and increasingly tied to AI anxiety and community concerns. The Iran war has not been a dominant driver of EV adoption; U.S. EV sales are down mainly because tax credits disappeared, while global sales continue to rise. Europe’s long-term strategic response to conflict is to reduce import dependence, increasing support for domestic or regional energy resources. Bloom Energy benefits from the current power-buildout cycle because operating-data-rich technologies and easier permitting matter when gas turbines are scarce. A big utility can’t easily isolate a county for special rates, but a small utility with a very large load addition could potentially use tariffs to spread benefits across all customers. Solar will likely keep growing strongly because it is fast to deploy, increasingly supported by batteries, and still easier to scale than gas in today’s supply chain environment. C&I solar will not scale unless both economics improve and AI/automation dramatically reduces soft costs and transaction friction. The grid and utilities are slower than behind-the-meter buildout, but once regulatory and planning processes are moving, they can deliver very large amounts of new capacity. AI-related power demand will accelerate both behind-the-meter generation and utility-side investment rather than one replacing the other.
Data Points: Data center capacity operated by Energy Hub: 3.4 gigawatts - Used in sponsor copy describing virtual power plants built from 2.5 million customer devices. Customer devices aggregated by Energy Hub: 2.5 million - Sponsor ad about VPPs turning customer devices into dispatchable grid capacity. Bloom Energy company size: ~$50 billion (previously ~$80 billion) - Khan notes Bloom is not an underdog despite being under-discussed. Potential tariff impact in small utility territory: 20% to 50% subsidy - Khan speculates large data-center loads could materially subsidize other customers’ rates in small utility systems. Example load/system ratio: 1 GW load on a 3 GW peak-load system - Illustrates the type of small-utility territory where a special tariff might work. Europe EV sales growth: +29% year over year - Khan cites Europe as still showing strong EV growth despite the Iran war and broader energy uncertainty. Global EV sales growth outside Europe and the U.S.: Doubling year over year - Shows global EV adoption remains strong outside the U.S. and Europe. U.S. EV sales change: -21% year over year - Khan attributes the decline mostly to the loss of tax credits, not the Iran war. Oil price level: around $100/barrel - Khan says prices rose but not to the extreme levels that would have dramatically shifted EV demand. Hypothetical oil price spike discussed early in war: $200/barrel - Referenced as an early fear that did not materialize. Data-center delay estimate: 6 months to 1 year - Khan says regulatory pushback can delay buildout, even if it won’t stop it. Solar forecast stance: bet the over - Khan argues historical forecasts repeatedly underestimate solar buildout. Time horizon for grid buildout: next 2 years - He expects more capacity to be built on the grid than in any year since before the turn of the century.
Pivotal Quotes: "the degree to which the data center backlash gained a foothold as quickly as it did" — Shail Khan: Explaining the biggest surprise in 2026 so far and why the backlash feels faster and broader than expected. "we need to pace the development of AI" — Referenced by Shail Khan (quoting Dario Amodei): Discussing the emerging AI backlash and the public debate over slowing AI/data-center expansion. "I think the market does not fully appreciate that, like those gears have really only gotten spun up over the past couple of years" — Shail Khan: Arguing that grid and utility investment will accelerate more than markets currently expect.
Implications: Expect more friction, delays, and local bargaining around data centers, but not a stop to AI infrastructure. Solar, fuel cells, and grid investment should keep growing, while utilities may become central to pricing, permitting, and community negotiations.