Future Inventions by 2100: 7 Technologies to Watch

Which future inventions could exist by 2100? Seven stand out because early versions already work today: health digital twins, direct air capture of carbon dioxide, brain-computer interfaces, brain-responsive virtual reality, cheap clean energy with long-lived storage, manufacturing at the scale of molecules, and autonomous vehicles. None is guaranteed. Each section below says what exists now and what would have to happen next.

Predictions 75 years out are mostly wrong in the details, so this list sticks to technologies with a working prototype or product in 2026. The interesting question for each is how far the prototype is from the version that would change daily life.

Digital “Twins” and Smart Wearables: Monitoring Health in Real-Time

A digital twin is a computer model of one specific patient, kept up to date with data from that patient. Smart wearables supply the data from embedded sensors. Together they could give patients and doctors continuous, personal health readings instead of a snapshot at an annual check-up.

By analysing vitals as they come in, a digital twin could flag an irregularity early, such as a heart arrhythmia, so that a patient seeks treatment before the condition becomes serious. Patients would see their own data too.

Market forecasts for these tools vary widely and change every year, so we do not quote them here. The direction is clearer than the numbers: by 2100, continuous monitoring could shift much of medicine from reacting to illness towards preventing it.

Consumer wearables can already flag irregular heart rhythms for follow-up by a doctor, and research groups are building computer models of individual patients’ hearts. Early intervention powered by biometric data is why digital healthcare makes this list.

Direct Air Capture: Pulling Carbon from the Sky

Direct air capture (DAC) removes carbon dioxide from the atmosphere. It uses chemical sorbents or solvents to pull CO2 molecules out of ambient air, and the captured carbon can then be stored underground or reused.

The IPCC’s 2022 mitigation report calls carbon dioxide removal unavoidable for reaching net zero, and in its modelled pathways that limit warming to 1.5 °C, direct air capture with storage removes anywhere from zero to 310 billion tonnes of CO2 between 2020 and 2100 (IPCC AR6 WGIII Summary for Policymakers, C.3.5 and C.11). Today’s plants remove thousands of tonnes a year, while the top of that range implies billions of tonnes a year, roughly a million times more.

Real-world DAC projects already show that the technology works. From 2017 to 2022 Climeworks’ first commercial plant, in Hinwil, Switzerland, sold captured CO2 to a nearby greenhouse. Orca in Iceland, opened in 2021 with a capacity of up to 4,000 tonnes of CO2 a year, sends the captured gas underground where it mineralises in basalt. Its larger neighbour Mammoth is designed for 36,000 tonnes a year (Nordic Investment Bank).

Cost and energy use are the obstacles. Pulling CO2 out of air, where it makes up about 0.04% of the gas, takes far more energy than capturing it from a smokestack, so DAC only helps if it runs on low-carbon power. By 2100 it could matter a great deal, but only alongside deep cuts in emissions, never as a substitute for them.

Brain-Computer Interfaces: Linking Minds and Machines

The boldest item on the list is a direct link between the brain and computers. Early versions exist. In 2023 a Stanford-led team reported in Nature that a woman who could no longer speak intelligibly because of ALS had her attempted speech decoded from implanted electrodes at 62 words per minute, with about one word in four wrong on a 125,000-word vocabulary (Willett et al., 2023). In January 2024 Neuralink implanted its first human patient, a man with tetraplegia who has since used it to move a computer cursor by thought.

These are medical devices for people with paralysis, implanted by surgeons, and they still need regular recalibration. A 2100 version would have to be safe enough for healthy people, last for decades in the body and carry far more information in both directions. If that happens, typing, searching and controlling machines could become as quick as thinking about them. Claims about uploading minds or learning skills instantly remain science fiction.

Hyper-Personalized VR: Unlocking the Full Potential of Our Minds

Virtual reality today is a headset that tracks your head, hands and eyes. Combine it with the brain-computer interfaces above and VR could respond to intention and attention directly, adapting a scene to what you are thinking about rather than to where you point a controller.

The most convincing uses are practical. A surgeon could rehearse a specific patient’s operation on that patient’s digital twin. A student could walk through a historical site or the inside of a cell. People with paralysis who already use implants could move and interact in virtual spaces that their bodies cannot reach.

The limits are the same as for brain implants: safety, bandwidth and the slow pace of medical approval. Full sensory immersion, a simulation indistinguishable from reality, would need a way to write signals into the brain at a precision no current device approaches.

The Clean Energy Revolution: Long-Lived Batteries and Cheap Renewables

The cost of solar power has already fallen faster than almost anyone predicted. The International Renewable Energy Agency puts the global average cost of electricity from new utility-scale solar plants at about $0.044 per kilowatt-hour in 2023, roughly 90% below 2010 (IRENA, Renewable Power Generation Costs in 2023, as reported by PV Tech). Cheaper solar panels and larger wind turbines keep pushing costs down.

Storage is the harder half. Solid-state and lithium-air batteries promise more energy per kilogram, though lithium-air cells still wear out after relatively few cycles. Exotic ideas such as nuclear diamond batteries can run for decades, but only at very low power.

Firm low-carbon power will matter too. Small nuclear reactors are one candidate. Further out, solar collectors in orbit could beam energy to receivers on Earth, an idea we examine in our explainer on space-based solar power.

By 2100 the likely change is a grid that runs mostly on renewables and storage, with fossil fuels a backup or gone. That would be an engineering achievement on the scale of electrification itself.

Molecular 3D Printing: Building Objects Atom by Atom

Molecular manufacturing would build objects by placing atoms and molecules exactly where a design says. Individual atoms have been moved on purpose since 1989, when IBM researchers Don Eigler and Erhard Schweizer used a scanning tunnelling microscope to arrange 35 xenon atoms into the company’s logo on a nickel surface cooled to about 4 kelvin (Nature, 1990).

The gap to a useful printer is enormous. A single gram of carbon holds about 5 × 10²² atoms, so placing them one at a time is hopeless; any real molecular printer would need billions of tools working in parallel, or chemistry that assembles itself. Materials engineered at the atomic scale already exist, such as the single-atom sheets described in our comparison of graphene and silicon.

If the problem is solved this century, the payoff would be materials with exactly the properties a designer wants, and medical treatments built to fit one patient’s cells.

The Age of Autonomous Vehicles: Revolutionizing Transit on Land, in the Skies and at Sea

Self-driving cars are no longer a prototype. Waymo opened fully driverless rides to the public in the Phoenix area in October 2020, and by March 2026 it operated robotaxis in ten US cities (TechCentral, March 2026). Delivery drones and autonomous ships are in trials.

By 2100, most vehicles on roads, in the air and at sea could drive themselves. That could make roads safer if the systems prove more reliable than human drivers, and cars that are summoned instead of owned could free much of the space cities now give to parking.

Not every transport dream survives contact with cost. Hyperloop One, which raised more than $450 million to build high-speed tube lines, shut down at the end of 2023 without winning a contract to build one (Fortune, December 2023). Autonomy is the safer bet, because it improves vehicles and infrastructure that already exist.

Conclusion: 7 Disruptive Innovations

All seven technologies already exist in an early form, which is why they are better bets for 2100 than flying cars or teleporters. The two with the clearest path are cheap clean energy and autonomous vehicles, where the prototype is already a product. The two with the longest road are brain-computer interfaces and molecular manufacturing, where the physics allows them but the engineering is decades away. Other candidates, such as quantum computing, could join the list as they mature.

Updated October 2026: rewritten to show where each technology stands today, with sources for brain-computer interfaces, solar costs, atom-scale manufacturing and self-driving vehicles.

@online{scienceshot-the-world-in-2100-2023,
  author       = {Turner, Alex},
  title        = {{Future Inventions by 2100: 7 Technologies to Watch}},
  organization = {ScienceShot},
  date         = {2023-07-13},
  year         = {2023},
  url          = {https://scienceshot.com/post/the-world-in-2100-7-innovations-set-to-transform-our-world}
}