The world of 2100 will be shaped by a handful of forces already in motion: artificial intelligence that likely surpasses human capability across most tasks, energy systems built around fusion and renewables, a global population that has already peaked and begun shrinking, and a climate that is measurably warmer regardless of what we do between now and then. The specifics depend on choices made in the next few decades, but the broad strokes are already visible in today’s research projections.
AI Will Likely Outperform Humans at Most Tasks
A survey of thousands of AI researchers, conducted by polling authors of recent AI publications, found that the median estimate gives a 50% chance of machines being able to accomplish every task better or more cheaply than humans by 2047. Some researchers placed that milestone within a few years; others put it hundreds of years out. But the central tendency points to human-level AI arriving well before 2100, not at the tail end of the century.
If that timeline holds, AI by 2100 won’t just be a tool people use. It will be the default way most cognitive work gets done: scientific research, engineering design, medical diagnosis, legal analysis, software development, creative production. The economic implications are staggering. Entire professional categories that exist today could be fully automated, and the nature of “work” for most people may look nothing like what we recognize now. The biggest open question isn’t whether AI reaches this level, but how societies restructure themselves around it, whether through universal income programs, new forms of human contribution, or something no one has proposed yet.
Robotics will follow a similar curve. Once AI can reason and plan at a human level, the main bottleneck for physical automation shifts to hardware engineering and manufacturing cost, both of which tend to fall over time. By 2100, autonomous systems will likely handle most physical labor, from construction and farming to eldercare and household maintenance.
Energy: Fusion, Renewables, and the End of Fossil Fuels
The U.S. Department of Energy has laid out a roadmap targeting commercial fusion power on the grid by the mid-2030s. That timeline is ambitious, and fusion has famously been “30 years away” for decades. But private investment in fusion startups has surged, and several companies are now building prototype reactors rather than just running experiments. Even if the first commercial plant slips to the 2040s or 2050s, fusion would still have half a century to scale before 2100.
Fusion produces energy by combining light atoms (typically hydrogen isotopes) rather than splitting heavy ones like conventional nuclear plants. It generates no long-lived radioactive waste and carries no risk of meltdown. If it works at commercial scale, it offers effectively unlimited clean energy. Combined with solar and wind, which are already the cheapest new electricity sources in most of the world, the energy landscape of 2100 will almost certainly be dominated by zero-carbon sources. Fossil fuels may still exist in niche applications, but their role as the backbone of civilization’s energy supply will be over.
Cheap, abundant clean energy changes everything downstream. Desalination becomes affordable enough to supply fresh water to arid regions. Carbon capture becomes economically viable at the scale needed. Energy-intensive manufacturing, from steel to concrete, can run without emissions. The constraint on human activity shifts from “how much energy can we produce?” to “what do we want to build?”
A Warmer Planet, No Matter What
Even under the most optimistic emissions scenario, global temperatures by 2100 will be 1.0°C to 1.8°C warmer than pre-industrial levels. Under a middle-of-the-road pathway where countries partially follow through on climate pledges, warming reaches 2.1°C to 3.5°C. In a worst-case scenario with continued heavy fossil fuel use, temperatures climb 3.3°C to 5.7°C.
Sea levels follow a similar pattern. The best-case projection puts the median rise at about 0.28 meters (roughly 11 inches) by 2100, while the worst case reaches a median of 0.68 meters (about 27 inches), with an upper range exceeding one meter. Even one meter of sea level rise reshapes coastlines worldwide, threatening cities like Miami, Jakarta, Shanghai, and Mumbai. Some island nations face partial or total submersion.
Technology will play a major role in adaptation. Carbon capture and storage, which currently removes only about 45 million tonnes of CO₂ per year globally (with just 9 million tonnes going into dedicated storage), needs to scale to hundreds of gigatons stored by century’s end to meet climate goals. Research published in Nature suggests this is physically feasible using growth models, but the gap between current capacity and what’s needed is enormous. Seawalls, floating architecture, managed retreat from coastlines, and heat-resistant infrastructure will all become standard features of life in 2100, particularly in vulnerable regions.
Fewer People, Older Populations
The world’s population is expected to peak in the mid-2080s at around 10.3 billion, then decline slightly to about 10.2 billion by 2100. That peak is 700 million people lower than what the UN projected just a decade ago, reflecting faster-than-expected fertility declines across much of Asia, Latin America, and parts of Africa.
A shrinking population changes the texture of daily life. Labor shortages become chronic in countries that are already aging rapidly, like Japan, South Korea, Italy, and China. This accelerates the adoption of AI and robotics out of necessity, not just efficiency. Healthcare systems will be oriented primarily around aging-related conditions. Cities in depopulating regions may contract, with entire neighborhoods emptied out while other urban centers continue to grow.
The combination of fewer young people and vastly more capable AI creates a world where economic output is less tied to human labor than at any point in history. Whether that translates into widespread prosperity or deepening inequality depends entirely on political and institutional choices made between now and then.
Medicine and the Human Body
AI-driven drug discovery is already compressing timelines that used to take a decade into a few years. By 2100, this process will likely be measured in weeks or months. Diseases that are currently managed but not cured, including most cancers, Alzheimer’s, and autoimmune conditions, will likely have effective treatments or cures. Gene editing tools descended from today’s CRISPR technology will be far more precise, potentially allowing doctors to correct genetic diseases before birth or even modify adult cells to reverse age-related damage.
The more speculative question is whether 2100 brings meaningful life extension. If AI accelerates biological research at the pace many researchers expect, it’s plausible that people born in the latter half of this century could live well past 100 as a routine matter, not as an outlier. This would compound the demographic shifts already underway, creating societies where four or five generations coexist simultaneously.
Daily Life and Connected Infrastructure
The physical environment of 2100 will be saturated with intelligence in a way that makes today’s “smart” devices look primitive. Buildings will generate their own energy, manage their own climate systems, and repair minor structural issues autonomously. Transportation in dense areas will be almost entirely autonomous, with human-driven vehicles restricted or obsolete. The distinction between “online” and “offline” will have dissolved completely, with digital interfaces embedded in surfaces, clothing, and possibly the body itself.
Food production will look radically different. Vertical farms, precision fermentation (using microorganisms to produce proteins, fats, and other nutrients), and lab-grown meat are all in early commercial stages today. By 2100, these technologies will have had 75 years to mature and scale. Traditional agriculture won’t disappear, but a large share of calories, particularly animal proteins, will come from controlled indoor environments that use a fraction of the land and water.
Space activity will expand significantly, though the science fiction vision of mass colonization remains unlikely by 2100. More realistic is a permanent human presence on the Moon, research outposts on Mars, and a thriving orbital economy built around manufacturing, tourism, and satellite infrastructure. The cost of reaching orbit has already dropped by orders of magnitude in the past two decades, and that trend shows no sign of stopping.
The Biggest Uncertainties
Projections for 2100 carry enormous uncertainty, and the biggest wildcards are not technological but political and social. Wars, pandemics, and governance failures can derail even the most promising technological trajectories. The concentration of AI capability in a small number of companies or governments could create power imbalances unlike anything in history. Climate tipping points, such as the collapse of major ice sheets or the release of methane from thawing permafrost, could push warming beyond any current model.
What’s most likely is that 2100 will be defined by sharp contrasts: regions that adapted early and invested in new technology thriving alongside regions that were locked out by poverty, conflict, or geography. The tools to build an extraordinary civilization will almost certainly exist. The question, as it has always been, is whether they’re deployed wisely and shared broadly enough to matter.