Electrification Is Everything

ELI5/TLDR

Michael Liebreich, who built the firm that became Bloomberg’s clean-energy data arm, gives a 40-minute tour of where the energy world actually is. His headline: nearly all new power being added to grids worldwide is now wind and solar, batteries are at the same tipping point solar hit a decade ago, and the real story isn’t generating clean power but electrifying everything that currently burns fuel — cars, heating, factories. He argues the standard “the transition has failed” line rests on a misleading statistic, and that the developing world is going electric not for green reasons but because it’s simply cheaper.

The Full Story

Two stories about the same crisis

Liebreich opens with the war in the Gulf and the two narratives it has spawned. One says: we took our eye off the ball, got obsessed with net zero, leaned on wind and solar that aren’t resilient, and should have secured reliable fossil supplies. The other says clean energy is the resilience, and that an oil-price shock is fundamentally a fossil problem that won’t have a fossil solution. His answer is “a little bit of both” — short-term you scramble to secure supply, but the durable fix is on the clean side.

His sharper point is about hidden costs. Critics love to say wind and solar only look cheap because someone else pays for the grid connections — the “externalities.” Liebreich flips it. Since the 1973 Arab oil embargo, the oil price has lurched 30% or more in a single year on fourteen separate occasions. Those shocks batter the whole economy and nobody pays for them through the pump price.

The Climate Change Committee in the UK says one more shock would actually cost the UK economy more than the extra cost of getting to net zero.

The great clean energy acceleration

He traces his own career as a measuring stick. When he founded New Energy Finance, global renewable investment was $36 billion a year. Add grid spending and it was maybe $230–240 billion. Today the clean-energy total is $2.3 trillion — roughly a 10x jump, some of it inflation.

The physical kit tells the same story. Of everything being added to electrical grids worldwide in 2025, somewhere around 90% is wind and solar. Wind and solar together now generate more than gas. Add hydro and renewables beat coal — “King coal.” And wind alone, and solar alone, each now out-generate nuclear. None of this was thinkable when his industry was still called “alternative energy.”

Batteries are where solar was in 2015

This is the part he’s most excited about. A 2019 forecast his old team made for battery growth looked aggressive at the time; reality has blown past it. The trick of exponential growth, he notes, is that today’s dramatic numbers will look like a flat little stub on the chart in a decade — the way the early years of solar now look trivial next to the curve.

What makes batteries special isn’t just cost, it’s location. Think of it like this: for as long as he’s worked in energy, storing serious amounts of power meant pumped hydro — water pumped uphill, released through turbines later. There was no real alternative. A battery, by contrast, you can put anywhere: a home, a factory, either side of a congested grid line.

You can put it in your home. You can put it in your factory… a battery is more valuable than pumped hydro. Never mind the cost. It’s more valuable.

In Australia — world champion at rooftop solar, now with 500,000 home batteries — this is already pushing gas out of setting the price for more and more hours, and electricity prices are actually falling. He calls batteries the next “turn of the screw”: zero-marginal-cost power that blows up how electricity is priced.

China plays chess, America plays air guitar

His most-quoted line of the talk:

China understood this. They’ve been playing chess. We’ve been playing draughts. America’s been playing air guitar.

China went from the world’s fifth-largest car exporter to first in three years and now dominates the entire clean-energy stack — solar, batteries, critical minerals. The US, he says, has chosen to become a “petrostate” while China becomes an “electrostate.”

Europe is stuck in the middle, hammered by high industrial electricity prices (worse in the UK), trying to manufacture against a brutal cost gap just as the technologies it makes are going electric. He reaches for Gramsci: “The old is dying and the new struggles to be born. In this interregnum, a great variety of morbid symptoms appear.”

The US is more complicated than it looks. EVs and offshore wind hit a wall, but tax breaks survived, and carmakers are repointing their battery factories at grid storage — enough that by 2027 the US could double all its electrical storage every two years. Clean energy there is wounded, not dead.

The statistics people use to lie to you

Two big ones. First, the AI data-center frenzy. Texas has a 400-gigawatt connection queue against maybe 70–80 GW of actual generation. His rule of thumb:

The way to read all stories about AI data centers is divide by 10… and you might have the right number.

Maybe divide by 20. He’s not denying AI is real — he uses it daily — just that the proposed buildout is fantasy, much of it pitched by real-estate developers who once built a 10-megawatt center and now claim they’ll build a gigawatt.

Second, and more important: primary energy. This is the stat behind every “the transition has failed, fossil fuels are still ~80% of energy” claim. Liebreich’s demolition is the intellectual core of the talk. Primary energy measures the raw fuel the energy industry burns — not what society actually uses. When you go from coal burning in a power plant to light an incandescent bulb, over to solar powering an LED, you cut primary energy by 95%, because the coal plant wastes most of its energy as heat. A fossil car to an electric car: 75% less primary energy. The combustion engine throws away roughly three-quarters of its fuel as heat; the electric motor doesn’t.

If somebody uses the word primary energy, please remember there’s one of two things happening. Either they don’t know what they’re talking about, or they do, but they hope you don’t.

So measuring the transition by primary-energy share is rigged against the clean side — clean tech wins partly by needing far less raw energy in the first place.

The model: it’s about growth, not targets

He contrasts two ways of thinking about the transition. The activist version starts from a fixed destination — “keep 1.5 alive” — and as emissions keep rising, demands an ever-steeper path until you’d need a time machine (which, he dryly notes, exists: it’s called carbon dioxide removal).

His version is the entrepreneur’s: don’t fixate on 2042, just grow. Energy is already 30% clean — we’re in the middle third, not the start. If real energy demand grows 2% a year while clean grows 5%, the small fast thing eventually eats the big slow thing. Fossil gets forced off the system by arithmetic.

The punchline, and a deliberate one: not “electrify everything” but “electrification is everything.” Road transport, space heating, and low-to-mid-temperature industrial heat can all electrify today — and that’s about 45% of global emissions, addressable now with electricity that’s getting cleaner everywhere. The hard stuff — aviation, cement, steel — shouldn’t be ignored, but pouring all the effort there means slow, expensive, shallow progress. China is electrifying its economy at 10% per decade, versus the 4% Europe and the US managed at their fastest.

And the “rest of world” — the 70% of humanity outside the US/China/Europe frame, where nearly all energy-demand growth lives — is going clean not out of ideology but cost. An eminent professor told him no one in Africa would buy an electric car; with a little help from Claude, Liebreich pulled the data: 60% of car sales in Ethiopia, 76% in Nepal. They go clean because it’s cheap, with just enough oil and gas to keep the lights on.

Key Takeaways

Primary energy is a rigged metric. It counts fuel the energy industry burns, not energy society uses. Coal-to-solar for lighting cuts primary energy ~95%; petrol-to-electric car cuts it ~75% — because combustion wastes most of its fuel as heat. “Fossil is still 80%” claims lean on this trick.
~90% of all new grid capacity added globally in 2025 is wind and solar. Each of wind and solar now individually out-generates nuclear; renewables beat coal; wind+solar beat gas.
Batteries are at solar’s 2015 moment — about to be as transformational as the wind/solar buildout itself. Their edge over pumped hydro is placement, not just cost: a battery goes anywhere, including either side of a grid bottleneck.
In Australia, batteries + rooftop solar are pushing gas out of price-setting, and consumer electricity prices are falling as a result.
“Divide by 10” rule for AI data-center announcements (maybe divide by 20). Texas’s 400 GW connection queue dwarfs its 70–80 GW of actual generation.
Clean investment grew ~10x in two decades — from ~$230–240B/year to $2.3 trillion.
The growth model beats the target model. Clean is already 30% of energy; if clean grows 5%/yr against 2% demand growth, fossil is squeezed off by compounding alone.
“Electrification is everything”: road transport + space heating + low/mid-temp industrial heat ≈ 45% of emissions and can electrify now. Hard-to-abate sectors (aviation, cement, steel) shouldn’t monopolize effort.
China electrifies at 10%/decade vs the West’s 4% at its fastest pace.
The “rest of world” goes clean on cost, not ideology — 60% of car sales in Ethiopia and 76% in Nepal are electric.
Nuclear is not a climate solution at scale: to merely catch up to wind+solar output you’d need ~1,250 Rolls-Royce SMRs or ~8,000 small modular reactors — orders of magnitude beyond any realistic buildout.
Oil prices have moved ±30% in a single year fourteen times since 1973 — fossil dependence is structurally volatile, and those shocks are an unpriced externality.

Claude’s Take

This is a strong talk and Liebreich is exactly the kind of source worth trusting carefully: he has the data lineage (he built what became BloombergNEF), he’s openly partisan toward clean energy, and he tells you so. The partisanship is the thing to keep in view. He’s a salesman for a worldview he genuinely believes, and the rhetorical moves are slick — the “petrostate vs electrostate vs state of confusion” framing, the Gramsci quote, “playing air guitar.” Memorable, and also designed to flatter people who already agree.

The primary-energy argument is the most valuable thing here and it’s genuinely correct physics. Combustion really does dump most of its energy as waste heat, so a like-for-like accounting of useful energy does favor electrification. Anyone who has heard “renewables have barely dented the 80% fossil share” should sit with this — it’s a real measurement problem, not spin. The “divide by 10” AI rule is a good corrective too, though it’s a vibe, not a model, and he knows it.

Where to stay skeptical: the 5%-clean-growth-eats-2%-demand model is arithmetically true but assumes the growth rate holds, and the last mile of any transition (the hard-to-abate sectors he waves at, plus the grid rewiring electrification demands) is where optimistic curves usually break. He’s also light on the integration problem — all that zero-marginal-cost intermittent power needs storage, transmission, and market redesign he gestures at but doesn’t price. And he picks his battery comparison points (“never mind the cost”) in a way that skates past the fact that cost still matters a great deal.

Eight out of ten. High signal-to-noise, several mental models worth keeping (primary energy, the growth model, divide-by-ten, batteries-as-placement), delivered by someone with real standing. It loses points for being a keynote rather than a debate — no one pushes back on the optimistic curves, and a 40-minute “skate across the surface” can’t defend its biggest claims. But it’s the rare energy talk that will change how you read a headline.