Why cost curves, electrification, and AI are ending legacy economics
A cliff, not a trend
In most major economies, new solar and onshore wind are now dramatically cheaper than new coal or gas — not by a few percentage points, but by margins so wide they break decades-old planning models.
Electric vehicles are already significantly cheaper to run, with near‑zero routine maintenance, and in many markets they are rapidly becoming cheaper to buy upfront than their internal‑combustion equivalents. Oil demand is not being debated away — it is being priced out, kilometre by kilometre.
Nuclear, increasingly offered as a hedge or delaying tactic, is losing ground on levelised cost to a combination of grid‑scale storage, home batteries, and vehicle‑to‑home (V2H). What was once framed as a baseload necessity is being undercut by systems that are faster to build, cheaper to scale, and infinitely more flexible.
At the same time, the electrified stack is moving beyond energy and transport and into daily life. In China, home robots and autonomous domestic systems are beginning to proliferate into households, not as curiosities, but as labour‑saving appliances powered by cheap electricity, batteries, and AI.
Once machines can see, move, learn, and work at near‑zero marginal energy cost, the implication is unavoidable: before most people realise what is happening, a large share of what we currently call “human jobs” will be automated away within a decade. Not because of ideology or policy, but because labour‑substituting systems become cheaper, faster, and more reliable.
This is demand pulled forward by abundance, not constrained by scarcity.
This is not a gentle slope.
It is a cliff.
Everything that follows flows from that single fact.
This isn’t a transition. It isn’t a debate. And it certainly isn’t a matter of opinion.
What we are living through is an existential, asymmetric disruption between two systems that cannot coexist. One survives only through inertia, scarcity, and narrative control. The other advances through collapsing cost curves, learning effects, and relentless deployment.
When systems realise they are losing, they don’t quietly adapt. They distort reality. They deny data. They repeat claims that are demonstrably false, louder and with greater confidence, daring their audiences to stop believing their own eyes.
That phase has arrived.
Every major disruption reaches it: the moment incumbents stop evolving and start lying — not because they are foolish, but because truth has become structurally inconvenient.
Denial as a survival strategy
The frequency and brazenness of public falsehoods today are not random. They are symptoms.
Claims like “China exports wind power but doesn’t even use it” — voiced publicly by figures such as Donald Trump, echoed by parts of legacy media, and recycled across incumbent‑aligned commentary — aren’t mistakes. They’re tells. They are not just misleading — they are objectively untrue. In 2025, wind supplied roughly 10.5% of China’s total electricity mix, more than double nuclear’s contribution.
That single data point matters because it exposes the broader pathology.
And it isn’t uniquely Chinese or American.
In deregulated power markets like Texas and Australia, a similar pattern plays out: lobbying for so‑called “capacity payments” to keep uneconomic gas plants online, or redefining “baseload” in ways that conveniently exclude renewables paired with storage. The language differs. The strategy is the same.
Reality has moved on. Narratives have not.
China is not deploying renewables for virtue signalling or geopolitical theatre. Nor are grid operators elsewhere pairing solar with batteries out of ideology. They are doing so because the mathematics leave no alternative. Wind, solar, batteries, EVs, and grid upgrades are not political projects — they are cost‑minimisation strategies executed at scale.
When something becomes cheaper, faster, and more reliable, it does not wait for consensus. It spreads.
Legacy systems understand this intuitively. That’s why denial has become a strategy rather than a mistake.
The architecture of the old world
Legacy systems were built for a different era.
They are optimised for scarcity, friction, and control. For chokepoints and rent extraction. For slow‑moving capital and long amortisation schedules. They depend on complexity not as a flaw, but as a feature — complexity that locks in dependence and justifies margin.
Oil pipelines. Centralised power plants. Fuel distribution networks. Manual maintenance regimes. Labour‑heavy processes. Opaque pricing. Regulatory lobbying dressed up as “reliability.”
This architecture only works in a world where alternatives are expensive and change is gradual.
That world is gone.
Cost curves: the real battlefield
Figure: Annual solar PV additions — reality vs forecasts — Forecasts assumed linear growth. Reality went vertical. Annual solar additions surged from ~252 GW (2022) to ~392 GW (2023), ~599 GW (2024), and ~760 GW in 2025, obliterating legacy planning models. (Source: EV Curve Futurist)
Figure: Global wind installations (2024) — Installed capacity and growth rates highlight how scale is compounding unevenly across regions, with China operating on a different deployment curve entirely. (Source: Visual Capitalist / Energy Institute)
Figure: China electricity mix (2025) — Wind at ~10.5% of generation, more than double nuclear, with low‑carbon sources reaching ~42% of total supply. (Source: lowcarbonpower.org)
The defining mistake of legacy thinking is believing this is a political fight.
It isn’t.
This is a contest between cost curves — and cost curves are ruthless.
Solar continues to undercut fossil generation. Wind scales past nuclear not because of ideology, but because of modularity and speed. Battery Energy Storage Systems flatten volatility, erase peak pricing, and displace gas peakers. EVs don’t need to replace oil overnight — they only need to nibble at demand until refinery economics break.
None of this requires persuasion.
It only requires repetition.
Quarter after quarter. Year after year.
Cost curves don’t argue. They don’t negotiate. They don’t compromise. They simply keep falling until the old economics snap.
Emotion is noise. The signal is in capital expenditure plans, utility filings, and quarterly reports.
Storage ends the last excuse
Figure: Grid‑scale battery deployment comparison — China’s December 2025 grid storage additions alone exceeded total U.S. installations for the entire year, underscoring deployment velocity rather than policy intent. (Source: Katusa Research / industry reports)
Intermittency was the final refuge of the legacy argument.
It is gone.
A gas peaker plant may take minutes to ramp. Grid‑scale batteries respond in milliseconds. Markets pay for speed, flexibility, and certainty — and the future is faster.
Once storage is layered onto wind and solar, the system stops being fragile and starts becoming anti‑fragile. Excess becomes input. Variability becomes advantage.
At that point, comparisons with legacy power aren’t ideological.
They’re embarrassing.
Demand doesn’t save legacy — it accelerates its end
Figure: Global EV sales growth (2016–2026E) — EV adoption accelerating from niche to ~25% of global car sales by 2025, with fleet size compounding ahead of expectations. (Source: EV Curve Futurist)
Table: EV growth data — Annual sales, market share, and cumulative fleet size illustrate how demand scales once cost parity is crossed. (Source: EV Curve Futurist)
There is a persistent misunderstanding: that rising electricity demand will rescue the old system.
It won’t.
As power gets cheaper, demand explodes. EVs, data centres, AI workloads, electrified heat, industrial processes, robotics — all pull electricity forward. This is not a problem for renewables.
It is their ultimate victory.
Rising demand forces investment into the cheapest, fastest‑to‑deploy supply. That is not coal, gas, or nuclear. It is solar, wind, and storage, built in parallel and scaled relentlessly.
Legacy systems see demand as a lifeline.
In reality, it accelerates their displacement.
Transport is crossing the cliff first
Figure: China new‑energy heavy‑duty truck sales (2025) — Monthly sales and penetration show the inflection, culminating in >50% share of new sales by December 2025. (Source: CV World / CnEVPost)
If this were still theoretical, heavy transport would be the last place it showed up.
Instead, it is the first.
In China, electric heavy‑duty trucks crossed the 50% share of new vehicle sales in December 2025. Not cars. Not buses. Freight — the most brutally economic, margin‑sensitive segment of transport. This happened decades earlier than even optimistic forecasts suggested just a few years ago.
That single datapoint destroys the idea that electrification is driven by subsidies or ideology. Fleet operators do not buy politics. They buy lower operating costs, higher uptime, and predictable energy pricing. When the economics flip, adoption doesn’t crawl — it snaps.
The same logic is now spilling into shipping.
Large vessels are being electrified not as prototypes, but as working assets. Tasmania‑built electric ferries from Incat are already operating, proving that even traditionally diesel‑locked segments are not immune once energy density, charging, and route economics align.
When freight and shipping move, the argument is over. These are not lifestyle purchases. They are balance‑sheet decisions.
And now, aviation is following.
Electric vertical take‑off and landing (eVTOL) aircraft are already operating as air‑taxi services in places like Dubai and China, with others close behind. What was once dismissed as science fiction is becoming a regulated, commercial reality — not because it is flashy, but because short‑haul, urban air mobility is brutally exposed to fuel costs, maintenance overheads, and noise constraints.
As with trucks and shipping, the economics flip first in the segments where utilisation is high, routes are predictable, and downtime is expensive. Recent gains in battery energy density, fast‑charge capability, and high‑cycle durability are what make this possible — short missions, rapid turnarounds, and high asset utilisation finally pencil.
Regulation follows economics. Once safety data accumulates and operating costs undercut helicopters and ground congestion, approval accelerates. The transition doesn’t wait for consensus.
It scales.
Bettrification: the compounding stack
Figure: Electrification vs Bettrification — Electrification replaces hardware; Bettrification rewrites systems, layering intelligence, autonomy, and optimisation on top of cheap electrons. (Source: EV Curve Futurist)
Figure: Lithium‑ion battery demand by sector (2025) — EVs and BESS dominate demand growth, with LFP chemistry expanding fastest due to cost and durability advantages. (Source: Benchmark Mineral Intelligence)
Energy is only the foundation.
This is Bettrification — the self‑reinforcing stack where electrification enables intelligence, and intelligence optimises electrification, creating a feedback loop of collapsing costs and expanding capability. It’s why costs don’t merely fall — they collapse in cascades, each layer reinforcing the next.
Layer on AI: forecasting demand, optimising dispatch, coordinating millions of distributed assets in real time. Layer on robotics: automating manufacturing, logistics, mining, construction, and maintenance. Add autonomous systems, software‑defined factories, and machine learning compressing design cycles.
This is not a linear transition.
It is a stacked acceleration.
Electrification enables batteries. Batteries enable storage. Storage enables renewables. Renewables power AI. AI accelerates automation. Automation collapses costs further.
Legacy systems are fighting a multi‑front war with single‑use weapons.
Rage replaces adaptation
Unable to compete on cost or speed, legacy systems turn to something else: emotion. Incumbents scream into the air — like Donald Trump raging against windmills — to little effect, mistaking volume for leverage and denial for strategy.
Rage at renewables. Rage at EVs. Rage at batteries, data centres, AI, and anything that smells like the future. Rage at China. Rage at physics itself, rebranded as culture war.
Rage is the political externalisation of negative margins.
Watch the rhetoric closely. Claims become louder, simpler, and more detached from reality. Falsehoods are repeated not because they convince skeptics, but because they reassure believers.
This isn’t confidence.
It’s panic.
Markets don’t care about identity
Here is the part that makes this irreversible.
Markets do not care about nostalgia. They do not care about identity. They do not care about how loudly something is defended.
They care about cost, reliability, and scale.
Once a new system is cheaper, capital flows. Once capital flows, talent follows. Once talent follows, the old system starves — slowly, then all at once.
Politics becomes theatre. Balance sheets decide the outcome.
What this means for you
For investors, capital does not get destroyed — it migrates. It flees high fixed‑cost, long‑life assets exposed to physics risk and flows toward modular, scalable systems riding learning curves. The danger is not volatility. The danger is being trapped in assets whose cash flows assume a world that no longer exists.
For leaders and businesses, assumptions about energy costs, supply chains, and competitive moats are already obsolete. If your strategy depends on energy being scarce, slow, or volatile, it is built on legacy economics.
For individuals, the dividing line is simple. Skills tied to control, friction, and gatekeeping are legacy. Skills tied to systems thinking, software, electrification, automation, and optimisation are future‑proof.
The question is no longer whether the shift happens.
It is where you stand when it compounds.
How legacy actually ends
There will be no dramatic collapse. No final showdown. No moment where the lights go out and everyone agrees the old world lost.
Legacy systems don’t fall.
They fade.
Project by project. Plant by plant. Contract by contract. Balance sheet by balance sheet.
Engineers stop choosing them. Investors stop funding them. Young companies stop building on them.
Not because they were defeated — but because they became irrelevant.
The future doesn’t argue
The future does not need to win debates.
It does not need to persuade legacy audiences.
It only needs to keep getting cheaper.
And it is.
The future doesn’t gloat.
It compounds.
The cliff has already been crossed. Build accordingly.
Further reading (EV Curve Futurist)
For readers who want deeper context across energy, materials, AI, and the broader Bettrification thesis, these companion essays expand the arguments made here:
- Materials for a Bettrified World — why batteries, copper, lithium, and critical materials sit at the heart of the new stack.
- Dawn of the Age of Bettrification — the foundational essay defining the electrification + intelligence + automation convergence.
- Energy Transition: My Predictions for 2026 — near‑term signals and inflection points to watch as cost curves bite.
- AI and Robotics: The New Industrial Wave — how intelligence and automation accelerate every layer of the transition.
- 2025 Energy Transition in Review: The World Repriced — how markets silently adjusted while narratives lagged.
- Solar Dominion: The Decade China Locked Down Energy and Matter — the scale and speed behind China’s energy build‑out.
- Debunking Wind Power Myths: Separating Fact from Fiction — a data‑driven dismantling of persistent legacy talking points.
References
- Lazard — Levelized Cost of Energy (LCOE) Analysis (latest edition)
- BloombergNEF — Energy Transition Investment & Battery Price Surveys
- International Energy Agency (IEA) — World Energy Outlook and electricity statistics
- Ember — Global Electricity Review
- National grid operators and utility filings cited throughout (China, Australia, US ERCOT)