For sixty years, nuclear fusion has been the energy world’s most promising disappointment. Always thirty years away.
So why did Google sign a $1 billion power-purchase agreement for electricity from a reactor that does not exist, will not be built for years, and may never be economic? Two breakthroughs explain why: magnets and algorithms. First, the physics.
Contents
What is Nuclear Fusion?
Fusion is the mirror image of the nuclear power we know. Instead of splitting heavy atoms, it joins light ones, usually hydrogen isotopes, into helium. A small amount of mass becomes energy, as Einstein’s E = mc² implies. Because c (the speed of light) is vast, even tiny losses release extraordinary power.
Fusion delivers far more energy than fission. It uses hydrogen rather than uranium, and its fuel – drawn from water – is effectively limitless. The challenge is doing on Earth what the sun does effortlessly: forcing atoms to fuse at around 100 million °C. At that temperature, matter becomes plasma – a wild, writhing soup of charged particles that no physical container can hold. Magnetic fields keep it in place; lose them, and the reaction stops. It can only work under strict conditions, so there is no runaway reaction to fear.
Magnets and Algorithms
Two breakthroughs – one magnetic, one digital – have that goal within sight.
Magnets: In 2021, MIT and Commonwealth Fusion Systems (CFS) built a 20-tesla superconducting magnet strong enough to shrink a fusion reactor to one-fortieth the size of its predecessors. That leap turned a government project into a commercial race. CFS is building a 400-megawatt plant in Virginia, aiming to deliver power in the early 2030s. Google has agreed to buy half its output. Eni has committed $1 billion. These are long-dated bets. But they are backed by hardware, not hope, and that distinction matters.
Control: Plasma at 100 million °C is not just hot; it is violent. Instabilities bloom in milliseconds, faster than any human can react. Enter the algorithms. AI controllers designed at Princeton and tested on Switzerland’s TCV tokamak can now suppress these disruptions in real time, adjusting magnetic fields with speed and precision that makes the difference between a functioning reactor and an expensive sculpture.
Money: Capital has a talent for disbelief until it sees a profit. Fusion has reached that moment. CFS has raised over $2 billion privately. China spends $1.5 billion a year on fusion, nearly twice America’s public budget, because Beijing understands what Washington still debates: energy abundance is power.
The US National Ignition Facility has repeatedly achieved net energy gain since 2022. France’s WEST tokamak has sustained plasma for 22 minutes. Australia’s own HB11 Energy is building a laser-fusion facility in Adelaide. A 2024 global survey of 2,100 scientists across 87 countries forecasts commercial fusion in the 2030s.
Not so Fast
Investors know that being early is the same as being wrong. Even the bullish models say fusion will arrive too late to help mid-century climate goals. If a first plant starts in 2032, it will take decades to scale. Solar adds 600 gigawatts of capacity each year; fusion will be lucky to reach four by 2040.
The materials are not yet up to the physics. Reactors burn deuterium and tritium: the first abundant, the second rare. Tritium must be bred inside the reactor from lithium, keeping the so-called breeding ratio above one. The equations work, but the engineering doesn’t – yet. Components must survive years of neutron bombardment, and early plants will lose money on a heroic scale.
The greater danger is complacency – the belief that fusion will ride in like the cavalry, just in time to save us. That faith drains momentum from what works now: solar, wind, storage, heat pumps, efficiency. A tonne of carbon avoided in 2026 is worth ten promised in 2036. Fusion cannot be an alibi.
So Why Care?
Because fusion changes the long game. It offers energy unbound by weather, geography or scarcity. A bathtub of seawater could power a lifetime. The reaction stops the moment confinement fails. There is nothing to melt down.
If fusion can match renewables on cost, it will unlock abundance of a kind the world has never known. Desalination could erase water scarcity. Transport could run on fuel split from seawater. Nations that once fought for oil might cooperate to build reactors instead. The constraints that shaped civilisation for millennia – land, fuel, scarcity – would fall away.
What this Means for Business Strategy
Disciplined optionality is the right posture for business.
Accelerate today’s decarbonisation. Companies modelling fusion’s arrival while ignoring current emissions are practising creative accounting. Cut carbon now using renewables, storage and efficiency. This decade needs no breakthroughs, only discipline.
Monitor fusion as a hedge. For data centres, steelworks, and chemical plants, fusion could upend unit economics. Track the leaders – CFS, Helion, General Fusion, Tokamak Energy – and note their milestones.
Stay flexible. Energy investments live for decades. Whether you are building data centres, mines, or manufacturing capacity, plan for grids that can absorb firm, zero-carbon baseload if it arrives, whether from fusion, geothermal or long-duration storage.
The Long Game
Fusion will not solve climate change by 2050. But it might end energy scarcity by 2070. Climate change is a century-long project. Technologies that arrive too late for the first battle may yet decide the war.
The sixty-year joke is over. The only question now is pace. Keep your hand on the renewables lever. Keep your eye on the star we are learning to bottle.
We’d love to hear your thoughts – email luke@bwdstrategic.com or message him on LinkedIn if you’d like to continue the conversation.
About the Author
Luke Heilbuth is CEO of strategy consultancy BWD Strategic.