First – unless perhaps the Earth is struck by an asteroid – temperatures will rise. How much is a matter of debate; but much depends on the behaviour of mankind and developing technologies that can halt – or even reverse – the quantities of greenhouse gases being created.

Second is the inexorable rise of global population – inexorable, that is, until towards the end of the century when it’s predicted to stabilise then recede. Projections from the United Nations (UN) are that the numbers, eight billion in 2023, will peak around 2084 at about 10.3 billion and slowly decline. That assumes a sustained decrease in the global average fertility rate from 2.5 births per woman from 2015–2020 to 1.8 by 2100.

Other projections calculate that successful education and family planning goals could result in a peak during 2060–2070. Most models conclude that all the predicted growth between 2020 and 2050 will come from less developed countries, more than half from sub-Saharan Africa. Nigeria, says the UN, could become the world’s third most populous country, and China’s population could shrink by half.

What does an oil major think? 
These are factors economists at Shell worked around when creating three scenarios – that is, routes the energy transition might take – leading up to 2100. (The third edition was published in January 2026). They try to make sense of where we are now – with 80% of global power coming from fossil fuels but where the capacity of renewables has grown tenfold in the past decade. Each analysis factors in trade wars, artificial intelligence (AI) and climate politics to paint distinct pictures of the world at the turn of the next century.

Shell stresses these are ‘not created by using a crystal ball… absolutely not predictions or expectations of what will happen’, adding that ‘all scenarios probably hold truth, but… are likely to be wrong in one way or another’.

The Archipelagos scenario imagines a world where countries retreat into themselves. Trade disputes flare up, mineral supplies become weaponised and governments prioritise what they can control domestically over what might be cheaper or cleaner from abroad. Coal ‘hangs around far longer than climate activists would like’.

While China would race ahead with solar deployment to cut dependence on imported oil and gas, elsewhere ‘the transition stutters’. Priorities would shift towards domestic resources and opportunism in trade, both of which ‘tend to favour fossil fuel use’. Carbon capture struggles to get off the ground as funding evaporates. By 2100, temperatures have climbed 2.5°C above pre-industrial levels and fossil fuels still account for a quarter of global energy.

Shell’s Surge scenario portrays a different story, one wherein tech companies become energy giants. Data centres devour electricity ‒ by 2050, roughly 8% of global supply ‒ forcing them to build their own renewable capacity. These firms, ‘flush with cash and expertise’ in supply chain management, start mass-producing solar panels, batteries and hydrogen equipment. AI boosts productivity across all industries.

Small modular nuclear reactors come of age in the 2040s, powering data centres, then factories and eventually container ships. Despite higher energy demand overall, this world hits net zero emissions by 2080. Temperature rises peak at 2.1°C in the 2080s then start edging down.

The Horizon scenario sketches out what meeting the COP21 Paris Agreement of 2015 actually requires. Net zero by 2050 demands retiring ‘perfectly functional’ power plants early, slapping high carbon prices on industry and banning petrol cars. Horizon also relies on capturing more than 7 Gt/y of CO₂ by 2050, compared with just 50mn tonnes today, a 140-fold increase. Governments would dictate how much meat people should eat and which flights they should not take. Even then, world temperatures overshoot by 1.5°C with emissions yet to peak.

India and Africa, currently home to 1.5 billion people, could have a population of five billion by 2100. But rather than follow China's ‘coal-heavy’ development path, they may instead leapfrog straight to solar and batteries. 

All three Shell scenarios bump into the scramble for key minerals: copper, lithium, nickel and cobalt.

What about other voices? 
Some of these hunches were echoed last month in an article in open-access journal Science Insights. It conjures a world shaped by forces already in motion: AI, energy systems built around fusion and renewables, a global population shrinking from its peak, and a hotter world regardless of what happens between now and 2100.

By then the physical environment of 2100 will be ‘saturated with intelligence’. Labour shortages in countries with ageing populations, such as Japan, South Korea, Italy and China, may hasten the adoption of AI and robotics ‘out of necessity, not just efficiency’. Buildings will generate their own energy, manage their own climate systems and repair minor structural issues autonomously.

Sea levels, at best, are expected to rise by around 11 inches and, in the gloomiest forecasts, by 27 inches, putting at risk cities such as Miami, Jakarta, Shanghai and Mumbai, and threatening island nations with submersion. Sea walls, floating architecture, a retreat from coastlines and heat-resistant infrastructure will become commonplace.

What about other sources of energy, like nuclear fusion? 
There is one possible cause for optimism: last October the US Department of Energy released a strategy aiming for commercial fusion power to be on the grid by the mid-2030s. The concept has been around a long time – research into the possibilities for fusion reactors began in the 1940s – so it seems ripe to come of age.

The attraction of fusion – a means of electric power generation from energy released by combining two light atomic nuclei to form a heavier nucleus – is that it wouldn’t generate long-lived radioactive waste or carry the risk of meltdown. If commercially viable, reactors could contribute enormously to weaning the world off fossil fuels, and to a cheap energy mix that might mean desalination becomes affordable enough to supply fresh water to arid regions worldwide; and makes carbon capture economically viable at the scale needed.

However, management consultant McKinsey’s latest peer into the future concludes that fossil fuels will retain ‘a large share’ of the energy mix beyond 2050, with demand likely to plateau between 2030 and 2035. Natural gas will be more widely used for electricity generation, as well as for final consumption, perhaps displacing higher-emission fuels. Coal use may persist at higher levels than the firm predicted in previous forecasts.

‘Crucial alternative fuels are unlikely to achieve broad adoption before 2040 unless mandated [given] the current emphasis on affordability,’ it says. ‘Green hydrogen and some other sustainable fuels may not be competitive with traditional fuels in the near term.’

‘However, local market dynamics will influence the uptake of clean technologies. Nuclear energy, geothermal, hydropower and storage technologies that include batteries and pumped-hydroelectric-energy storage are likely to expand.’

What about reducing energy demand? 
Constructors of the environment have a role to play. Last year, a report from the International Energy Agency (IEA) underlined how urban design can impact demand for energy-hungry cooling systems – by up to 80% if buildings are properly insulated and given exterior shading. Integrating green spaces – parks and trees – into planning can reduce heat island effects and help cities cool down overnight.

The report points out that over 80 countries have mandatory building energy codes, and more than 100 minimum energy performance standards for air conditioning (AC) units. But standards and enforcement varies widely – the IEA says that globally, the average new AC unit sold is only about half as efficient as the best models available.  

Economic development is increasing access to AC in many countries, especially in emerging and developing economies, where more than 80% of projected electricity demand for cooling by 2050 is expected to occur. Currently, says the IEA, about 3.5 billion people live in hot regions, yet only about 15% own an AC unit.

That, inevitably, will change – the IEA predicts that across Southeast Asia the market for AC units will increase up to ninefold by 2040. But that threatens to put enormous strain on energy systems at times of peak demand. The IEA analysis found that if recent heatwaves continue in India – with potential temperature rises of more than 4°C – the extra peak load would be 47 GW.

Wilful tampering with – or neglect of – the environment remains the greatest threat to everyone. ‘The biggest wildcards are not technological but political and social,’ says Science Insights; and current events bear that out. Wars, pandemics and rogue governments are forever in the wings, able to ‘derail even the most promising technological trajectories’.

Optimists for the future will point to an energy system driven by the gathering momentum of renewable sources, some yet to be fully tapped; and the possibilities of these being hastened by AI. Set against that are the fears of dystopia: AI capability in the grip of a few all-powerful companies; or the doomsday prospect of ice sheet collapse and methane release from thawing permafrost heating up the world faster than anyone feared. While the brains and technology exist to create a clean, cheap energy system by 2100 for the benefit of all, the battleground – as ever – will be one of competing wills and vested interests.

• Further reading: 'The new energy trilemma – why national security is reshaping the global energy transition’. The energy transition is unfolding in a far more complex strategic environment than first envisaged, in which energy systems are no longer viewed simply as economic infrastructure, but are increasingly understood as strategic national assets – critical to economic resilience, industrial competitiveness and national security.
• ‘Why the grid will decide the UK’s energy future’. The UK’s energy transition is gathering momentum, driven by rapid electrification across transport, heat, industry and digital infrastructure. However, while much of the national conversation focuses on generation targets, it is the grid itself that will determine how quickly, equitably and productively the UK can reach net zero.