Strategy Consultant | Tech writer at http://ThePourquoiPas.com | AI writer of the Year | http://my.bio/adrienbook
I recently shared an article called “The “Next Big Thing” in Technology : 20 Inventions That Will Change the World”, which got a few dozen thousand hits in the past couple of weeks. This calls for a sequel. The previous 20 technologies were specifically centered on the next 20 years of technology development; but there’s a lot more to unravel when looking beyond the near future, though certainty obviously decreases with time.
Below are 20 technologies that will change the world by 2050 and beyond. This date and predictions are understandably vague and arbitrary, and we all know that predictions often fall flat (check my 2020 tech predictions if you don’t believe me). Regardless, the knowledge gained through planning for potential technologies is crucial to the selection of appropriate actions as future events unfold. Above all, articles such as this one act as catalysts to steer the conversation in the right direction
Life is far, far more complex than any of the technologies humanity has ever created. As such, it could make sense to use life’s building blocks to create an entirely new type of computational power. Indeed, for all the talks of Artificial Intelligence, nothing beats our mushy insides when it comes to learning and making inferences.
DNA computing is the idea that we can use biology instead of silicon to solve complex problems. As a DNA strand links to another, it creates a reaction which modifies another short DNA sequence. Action. Reaction. It’s not a silly idea : most of our computers are built to reflect the very organic way humans think (how else would we grasp computer’s inputs and outputs).
Humanity is pretty far from anything usable right now: we’ve only been able to create the most basic Turing machine, which entails creating a set of rules, feeding an input, and getting a specific output based on the defined rules. In real term… well, we managed to play tic-tac-toe with DNA, which is both dispiriting and amazing.
More on DNA Computing here [Encyclopaedia Britannica].
Smart dust is a swarm of incredibly tiny sensors which would gather huge amounts of information (light, vibrations, temperature…) over a large area. Such systems can theoretically be put in communication with a server or with analysis systems via a wireless computer network to transmit said information.
Potential applications include detection of corrosion in aging pipes before they leak (for example in drinking water… oh, hi Flint), tracking mass movements in cities or even monitoring climate change over large areas.
Some of the issues with this technology is the ecological harm these sensors could cause, as well as their potential for being used for unethical behavior. We are also far from something that could be implemented in the near future : it’s very hard to communicate with something this small, and Smart Dust would likely be vulnerable to environmental conditions and microwaves.
More on Smart Dust here [WSJ].
The name 4D printing can lead to confusion: I am not implying that humanity will be able to create and access another dimension (Only Rubik can do that). Put simply, a 4D-printed product is a 3D-printed object which can change properties when a specific stimulus is applied (submerged underwater, heated, shaken, not stirred…).
The applications are still being discussed, but some very promising industries include healthcare (pills that activate only if the body reaches a certain temperature), fashion (clothes that become tighter in cold temperature), and home-making (furniture that becomes rigid under a certain stimulus).
The key challenges of this technology is obviously finding the relevant components for all types of uses. Some work is being done in this space, but we’re not even close to being customer-ready, having yet to master reversible changes of certain materials.
More on 4D Printing here [Wikipedia].
Now THIS is real SciFi. Taking a page from biology, physics, mathematics, computer science and electronic engineering, neuromorphic engineering aims to create hardware which copies the neurons in their response to sensory inputs. Whereas DNA Computing aims to recreate computers with organic matter, Neuromorphic Hardware aims to recreate neurons and synapses using silicon.
This is especially relevant as we’re seeing an end to the exponential computing power growth predicted by Moore’s law (that’s quantum mechanics for you), and have to find new ways to calculate a bunch of things very quickly.
We’re not really sure how far this idea can be taken, but exploring it is, if anything, great for theoretical AI research. Should said research go further and become actionable, you’ll find me knocking on Sarah Connor’s door.
More on Neuromorphic Hardware here [Towards Data Science].
3D printing is still a solution looking for a problem. That’s partly because 3D printers are still too expensive for the average Joe, and not sophisticated and quick enough for large-scale manufacturing companies.
This may change over the next few decades: researchers have developed a method that uses a laser to ensure that incredibly tiny structures can be 3D-printed much, much faster (X1,000), while still ensuring a good quality of build. This method is called “femtosecond projection TPL”, but I much prefer “Nanoscale 3D printing” (because I’m a technological peasant).
Use cases are currently centered around flexible electronics and micro-optics, but quick discoveries around materials (both liquid and solid) leads researchers to think that they will be able to build small but imagination-baffling structures in the near future. One might imagine the medical community could use something like this…
More on Nanoscale 3D Printing here [Future Timeline].
As opposed to some of the other techs discussed in this article, this technology may not affect you directly, and is already being implemented (and will continue to be for a long, long time). Essentially, digital twins integrate artificial intelligence, machine learning and software analytics to create a digital replica of physical assets that updates and changes as its physical counterpart changes. Digital twins provide a variety of information throughout an object’s life cycle, and can even help when testing new functionalities of a physical object.
With an estimated 35 billion connected objects being installed by 2021, digital twins will exist for billions of things in the near future, if only for the potential billions of dollars of savings in maintenance and repair (that’s a lot of billions). Look out for big news on the matter coming out if the manufacturing, automotive and healthcare industries.
Why would I mention this ever-present idea as a technology to look out for in 2050? Easy : though we are talking about objects now, the future of digital twins rests in the creation of connected cities, and connected humans.
More on Digital Twins here [Forbes].
If one cuts through the blah blah (of which there is too much in this space), volumetric displays are essentially holograms. There are currently 3 techniques to create holograms, none of which are very impressive: illuminating spinning surfaces (first seen in 1948), illuminating gases (first seen in 1914), or illuminating particles in the air (first seen in 2004).
The use of volumetric displays in advertising (the primary focus for this concept) may be either greatly entertaining, or absolutely terrible because of potential impracticalities. You can imagine this easily by watching Blade Runner 2049).
I’m also dubious about the tech’s importance: computers were supposed to kill paper and I still print every single presentation I receive to read it. I don’t see hologram being anything else than a hype-tech attached to other more interesting techs (such as adaptive projectors).
More on Volumetric Displays here [Optics & Photonics News].
A brain-computer interface, sometimes called a neural-control interface, mind-machine interface, direct neural interface, or brain–machine interface, is a direct communication pathway between an enhanced or wired brain and an external device (If you start reading words like ElectroEncephaloGraphy, you’ve gone too far into the literature). If that sounds like something you’ve heard a lot about recently, it might have a lot to do with Elon Musk and a pig of his...
Beyond obvious and needed work in the prosthetic space, it’s the medical aspect which would be most transformative. A chip implemented in the brain could help prevent motion sickness, detect and diagnose cancer cells, and help with the rehabilitation of stroke victims. It could also be used for marketing, entertainment and education purposes.
But let’s not get ahead of ourselves : there are currently dozens, if not hundreds of technical challenges to wrestle with before getting anywhere near something the average person could use. First and foremost, we’d need to find the right material that would not corrode and/or hurt the brain after a few weeks, and get a better understanding of how the brain ACTUALLY works.
More on brain-computer interface here [The Economist].
Privacy: ever heard of it? Computer scientists are perfecting a cryptographic tool for proving something without revealing the information underlying the proof. It sounds incredible but not impossible once you wrap your head around the concept and the fact that it’s a bit more complex than saying “c’mon bro, you know I’m good for it”.
Allow me to simplify : Bob has a blind friend named Alice and two marbles of different colors, which are identical in shape and size. Alice puts them behind her back and shows one to Bob. She then does it again, either changing the marble or showing the same one again, asking if this is the same as the marble first shown. If Bob were guessing whether it was the same or not, he would have a 50/50 chance of getting it right, so she does it again. And again. And because Bob sees colors, he gets it right each time, and the chance that he guessed lucky diminishes. That way, Alice knows that Bob knows which marble is the original shown (and its color), without her ever knowing the color of any of the marbles. Boom, zero-knowledge proof. ZNP is this concept, applied to digitally to complex algorithms.
It’s easy to come up with VERY cool use cases. For example, if an app needs to know that you have enough money to put a transaction through : your bank could communicate that yes, that is the case, without giving an amount. It could also help identify a person without a birth certificate, or allow someone to enter a restricted website without needing to display their date of birth. Yay for privacy.
More on zero-knowledge proof here [Wired].
This one is easier to grasp as it has been part of the collective imagination for dozens, if not hundreds of years. Cars. But they fly.
Obviously, there are a lot of issues with this very sci-fi idea. We’re already struggling to stop people from attacking “classical” autonomous cars, so the jury is still out on whether it will ever come to be. Another issue is the fact that much of our world is built for traditional cars. Roads, buildings, parkings, insurance, licenses… everything would need to be destroyed and remade. It is likely that such cars will never see the light of day unless society crumbles and is rebuilt (2020’s not over yet).
There are currently 15 prototypes in development. I’d bet that none of these will ever come to light, except as playthings for the uber-rich. But hey, who doesn’t want to see Zuckerberg live out his mid-life crisis in the skies?
More on Flying Autonomous Vehicles here [Forbes].
This has also been a staple of SciFi for many years, for obvious reasons: imagine mixing robotics with enough Artificial Intelligence to entertain the idea of the digital world becoming physical. Welcome to your tape.
Before any of this can ever happen, we will need to improve robotics (robots don’t move so good right now) and create a new branch of AI research to explore a myriad of reactions such a technology would require to be operational. AMRs will also need nice, strong batteries, hence the current research into Lithium–silicon technologies.
Though no terminators are in sight, we’re starting to see such autonomous robots in warehouses, where they pick your Amazon purchase, and in the street, where they’ve begun bringing us our groceries.
More on Smart Robots here [EDx].
As I’ve mentioned in my previous articles, quantum computing will allow us to take leaps in the number of calculations a computer can do per second. A by-product of this is the fact that no password will be safe in a quantum world, as it should become possible to try all possible text and number combinations in record time.
Modern problems require modern solutions. Researchers at the Delft University of Technology in the Netherlands are working on a quantum Internet infrastructure where communications are coded in the form of qubits and entangled in photons (yes, light) flowing in optical fibers, so as to render them impossible to decrypt without disturbing the network.
In everyday words, that means that anyone listening in or hacking the network would disrupt the communication, rendering it unintelligible — data in such a state is, by nature, impossible to observe without altering it. The underlying science is fascinating, and I strongly recommend clicking on the link below to explore it.
More on Secure Quantum Internet here [Harvard School of Engineering].
This is yet another technology which is burgeoning today, but has yet a long way to go. At its heart, hyper-personalised medicine is genetic medicine designed for a single patient, making it possible to treat diseases that were once incurable, or that were too rare to be worth curing.
In 2019, a young girl named Mila Makovec, suffering from a rare and fatal genetic brain disease, was offered a tailor-made treatment (named Milasen — cute) to restore the function of the failed gene. Though she is not cured, her condition has stabilized, which is a big win.
The development of such personalized drugs is made possible by rapid advances in sequencing and genetic editing : creating a complete human genome sequence has gone from costing $20 million or so in 2006 to less than $500 in 2020. However, creating a drug still requires major resources (a year of development in Mila's case) and the mobilization of specialized teams. The question of cost therefore risks limiting the generalization of such treatments.
More on Hyper-Personalised Medicine here [MIT Technology Review].
Bioprinting is the process of creating cellular structures using 3D printing techniques, where cell functions are retained throughout the printing process. Generally, 3D bioprinting uses a layer-by-layer printing method to deposit materials sometimes referred to as bio-inks to create natural biological tissue-like structures which are then used in the fields of medical engineering.
A number of challenges pave the road ahead : we don’t know enough about the human body to implement these techniques safely, the price is very high, cells don’t live for very long when printed… the list goes on. And that’s without mentioning all the ethical questions such a technology raises.
There are nevertheless so many potential use cases that it’s well worth solving these issues : beyond allowing us to do transplants on amputees, it could also help us create “meatless meat”, leading to a more humane and more ecological meat industry.
More on Bioprinting here [Explaining the Future].
Several treatments intended to slow or reverse aging are currently in their testing phase. They block the aging of cells linked to age and reduce the inflammation responsible for the accumulation of toxic substances or degenerative pathologies, such as Alzheimer’s, cancer or cardiovascular diseases. In short, we’re not trying to “cure aging”, but instead seek to improve immune functions in older people.
Many studies are ongoing. In June 2019, the American start-up Unity Biotechnology, for example, launched a knee arthritis drug test. The biotech Alkahest, on the other hand, promises to curb cognitive loss by injecting young blood components.
Finally, researchers have been testing rapamycin, an immunosuppressant, as an anti-aging treatment for many, many years. The latter shows great promises, as it improves immune functions by as much as 20%.
The barriers are many : beyond the scientific costs, political pressure will need to be applied to key players to change the rules of healthcare as we know it. And we know how THAT usually plays out…
More on Anti-Aging Drugs here [University of Michigan].
Because of AI’s complexity, the computing power required to train artificial intelligence algorithms and create breakthroughs doubles every 3.4 months. In addition, the computers dedicated to these programs require a gigantic consumption of energy.
The digital giants are now working to miniaturize AI technology to make it accessible to the general public. Google Assistant and Siri thus integrate voice recognition systems holding onto a smartphone chip. AI is also used in digital cameras, capable of automatically retouching a photo by removing an annoying detail or improving the contrast, for example. Localized AI is better for privacy and would remove any latency in the transfer of information.
Obviously, because this space is ever-evolving, it is very difficult to see beyond the next few years of evolution — all we know is that many technical difficulties are still in the way (mathematically, mechanically, spiritually…).
More on Miniature AI here [MIT Technology Review].
The fact that Elon Musk makes a second appearance on this list is a testament to his very specific brand of genius. His hyperloop project consists of an underground low-pressure-tube in which capsules transporting passengers and/or goods move. Beyond removing air from the tube, friction on the ground is also removed, as the capsules are lifted by an electromagnetic lift systems. The capsules are propelled by a magnetic field created by linear induction motors placed at regular intervals inside the tubes.
Removing air and ground fiction would allow such a transportation method to reach insane speed : 1,102 km/h versus 885 km/h for planes at full speed (and the hyperloop can reach its top speed much faster than a plane). Other benefits include reduced pollution and noise.
However, this technology would require the creation of extensive tunnels, sometimes under cities. The price is fairly prohibitive : $75M per kilometer built. Other issues include making a perfectly straight tunnel, removing ALL air from the tube, and reaching the passengers in case of accidents. This has led some transportation experts to claim that the hyperloop has no future. Regardless, the memes are hilarious.
More on Hyperloop here [Tesla].
Asteroids have huge mineral resources. The idea behind space mining is that we could catch these asteroids, extract the minerals (especially the rare ones!), and bring them back to earth to sell them. Planets are also considered relevant in this discussion.
How hard could it be to make a lot of money from the final frontier ? Turns out, it’s fairly complex. Difficulties include the high cost of space exploration, difficulties finding the right asteroids, and the difficulty of landing on it when it’s moving at high speed (18,000 km/h on average). That’s a lot of difficulties. And that’s without discussing the potential trade and space wars that could result from two nations or companies having their eyes on the same space rock. So far, only the US and… Luxembourg (?) have passed laws in that regard.
However, if resources on earth become scarily scarce, and recycling is not an option, it might just become worth it.
More on Space Mining here [Financial Time].
An orbital solar power station, solar power satellite or space solar power plant would be an artificial satellite built in high orbit that would use microwave or laser power transmission to send solar energy to a very large antenna on Earth. That energy could then be used instead of conventional and polluting energy sources.
The advantage of placing a solar power plant in orbit is that it would not be affected by day-night cycles, weather and seasons, due to its constant “view” of the Sun.
This idea has been around since 1968, but we’ve still got a long way to go. Construction costs are very high, and the technology will not be able to compete with current energy sources unless a way is discovered to reduce the cost of launches (this is where Elon shines again). We could alternatively develop a space industry to build this type of power plant from materials taken from other planets or low gravity asteroids. Alternatively, we could just stop polluting the world, take one for future generation, and switch to less convenient / more expensive sources of energy...
More on Orbital Solar Power here [Forbes].
Honestly, I don’t know enough science to explain this one properly. But I’ll do my best! Teleportation, or the science of disappearing in one place to immediately reappear in another, is something that’s been in the popular imagination for decades now. We’re discussing something a bit simpler here: quantum teleportation is able to share information near instantaneously from one point to another, not matter.
We’re not talking about silly fish & chips recipe type of information, we’re talking about the make-up of entire molecules. In the early 2000s, scientists were able to transfer particles of light (with zero mass) over short distances. Further experiments in quantum entanglement led to successful teleportation of the first complete atom. This was followed by the first molecules, consisting of multiple atoms.
Logically, then, we could expect the first complex organic molecules such as DNA and proteins to be teleported by 2050.
I have no idea what to do with this information.
More on Quantum teleportation here [Wikipedia].
Technology has a tendency to hold a dark mirror to society, reflecting both what’s great and evil about its makers. It’s important to remember that technology is often value-neutral : it’s what we do with it day in, day out that defines whether or not we are dealing with the “next big thing”.
Good luck out there.
Create your free account to unlock your custom reading experience.