Predictions for the next decade: Hopes, dreams and reality checks: Part 6 of Tech That Reshaped our Lives in the Last Decade

This is part 6 of a 6-part series. Read Part 1 for Background. Go back to Part 5 on Hype Bubble of the 2010s.

Photo by Drew Beamer on Unsplash

At the end of the next decade, the tech in our lives will be faster, cheaper, more energy-efficient, better connected, and smarter than it is today. But all these advances are going to be much more boring than we imagine. The world of tech will look more the same than different. I predict:

1. No fully autonomous self-driving cars
2. No general-purpose robots for home or garden
3. No tablets fully replacing laptops or laptops replacing desktops or ereaders replacing books
4. No blockchain powered national currencies
5. Still no smart fridges that know what’s in them and automatically order food for us

But on the plus side, we can probably expect:

1. More ubiquitous and cheaper connectivity to a more compartmentalised internet
2. Longer-lasting batteries in smartphones and laptops
3. Better smartwatches that don’t have to be charged every day
4. Smarter cars that will park themselves reliably
5. Smarter homes that will manage heating, lights and key appliances efficiently and reliably

What the world needs now: Batteries

What the world needs now is more efficient energy storage. If there is one thing that could completely transform the world we live in it is a battery (be it in the form of a fuel cell, supercapacitor or something else) that can power a smartphone or a laptop for a week, fits in small spaces, charges almost instantaneously, does not explode easily, can be manufactured cheaply, uses abundant materials and can be recharged tens of thousands of times. Above all, it would make carbon reduction so much easier — solar and wind energy could be used more readily, electric cars would have fewer tradeoffs, etc. But in the world of tech, the windows for innovation in many areas would just explode open: wearable tech for one — smart watches, smart glasses, smart rings, … All of those things are available today but their usefulness is hampered by short battery life. What good are smart glasses that only last an hour?

My prediction is that we will not have such battery tech by the end of the next decade. I fervently hope to be wrong. But more likely, energy storage has the same in store for us as any other industrially mature technology from cars or aeroplanes to computer microprocessors. Boring, incremental improvements along various axes — not all of them noticeable. But incremental improvements add up — we can now get more out of our batteries that at the start of the decade — even if much of that improvement came from more power-efficient processors and other energy saving tricks. People like Elon Musk are betting on batteries improving by about 7% a year. That means that battery capacity would almost double by 2030 from what it is today. But there is no law here. We could make bigger leaps or grind to a halt. But if I were a billionaire, I’d fund open battery research alongside malaria and EMP-readiness.

5G and more accessible mobile networking

In 2001, I spent a year in Prague when the GPRS (2G) on my phone was my only internet connection. It was fine for email and the occasional web-search — not much of a downgrade on a dial-up. Then, there was a long period when the landline was far superior to any mobile connection. LTE (or 4G) changed all that. When the signal is plentiful and the needs are typical, you may not even notice you’re not connected to your local WiFi. What then can we expect from the future? Lower prices and unlimited data caps would be a good start — if not a certainty.

When it comes to a more connected future, 5G is getting a lot of press. But it probably won’t be until much closer to 2030 before we can tell whether it’s had the impact people are expecting. I don’t have any strong feelings about this either way but I’d certainly not say no to more devices always connected to the internet and each other in more places with less latency. If 5G could replace local broadband and we could get always connected laptops, I’d say that’s a future worth having.

LED Monitors and the foldable display

The 2010s were a time of display evolution. But they evolved into being pretty awesome. While 3D failed and plasma died a quiet death, high resolution (4K) and deep colour (HDR) displays became the norm on TVs, phones and even computer screens. But there’s also some tech on the horizon that hints at more revolutionary changes than that.

Imagine a world where almost any surface can be a screen. Any bit of a wall and even a roll up screen. Two different technologies are making their first appearance that may make this happen. They are LED displays and foldable displays.

2019 saw its first foldable phone to wide acclaim of potential but flexible displays have been making slow inroads for a while. Foldable tablets and more are being announced left and right — will we ever get to a place where we can just take a think display and roll it up like a newspaper? Possibly not but being able to fold a large iPad in half would be more than enough. I can certainly see that as a possibility by 2030. The underlying tech works in principle but still needs some engineering to be completely ready for the mass market. 10 years sounds like a reasonable time-frame to make something of it.

Much less flashy but with no less potential is the MicroLED display. Almost any new TV will use LED as its light source. But the image instead shows up on a thin layer of transistors (or something like that) through which the light shines. MicroLED displays, on the other hand, don’t have a single source of light. Instead, they consist of a lot of tiny LED lights that can display lots of colours. That may not sound that exciting, because display tech changes all the time, but what this enables is to have much smaller thin display squares that can be seamlessly combined into huge surfaces of any shape — even discontinuous. They will consume less energy, weigh less, and each segment can receive different signal at the hardware level. At the moment, they’re still more expensive than traditional displays and need to overcome other difficulties but the potential for having a corner display that follows the shape of a corner can open up new ways of designing interactive public spaces.

Struggle for control and regulation

“Information wants to be free” was the rallying cry of the Internet pioneers in the 1980s. The cluetrain manifesto in the 1990s talked of markets as conversations and about the time of the Arab Spring, the Internet was seen as a vehicle of ever expanding freedom. But at the same time, the unregulated nature of the internet was running into conflict with the regulated nature of the world, the internet was more and more interwoven with.

China started building the Great Firewall in the late 1990s but it was only in the 2010s that it became more pervasive and impactful. Many governments in the Middle East also started experimenting with various controls and limits on expression. Russia has also started asserting more control over what goes on in its virtual borders. This is not necessarily only related to online activities, both Russia and China are also developing their own operating systems that could run on the computers they want to control. Russia decided to kick off the new decade by a successful test of RUnet — a version of the internet disconnected from the rest of the world.

This is not limited to countries with a reputation for more central control of individual freedoms. The US government is waging a battle over encryption with providers of communication platforms, as is Australia and UK. The EU implemented regulations like the ‘right to be forgotten’ and GDPR which ostensibly protect the individual but limit the internet. At the same time, it is promoting new copyright regulations that would severely limit the ability to share information online. All of these efforts were met with opposition — sometimes more, at other times less successful.

Based on all this, I feel pretty secure in making the prediction that the next decade will be defined by more efforts by governments to assert more control over the internet. Be it regulation of companies like Facebook or Google or efforts to install backdoors in encryption. And it is also safe to assume that new technologies will continue to be developed to subvert this effort. Sometimes this effort will be laughably doomed to failure like the UK’s effort to limit access to pornography. Sometimes, it will be more sophisticated such as the Chinese efforts. VPNs will almost certainly continue to boom in the 2020s and will be coming up with more and more inventive ways of avoiding detection. My money is on the governments being more (but not completely) successful in their efforts for control but also that the consequences of these efforts will be less dire than predicted. (Although, they may be pointlessly annoying like the stupid EU cookie law.)

Adventures in automation

So far, The Fourth Industrial Revolution (4IR) has been a bigger business for the gurus than anybody else. True, more and more modern production factories are clean, quiet places — far removed from the noise and dirt I remember when I did my work experience stints in a car parts manufacturing machining department in the 1980s. It takes many fewer people to build a car than it did a few decades ago. But much of this is due to process refinement that makes the people involved more productive and assisted by clever, single-purpose machines. There’s no such thing as a general-purpose robot arm (let alone a robot) or general purpose intelligent software that could manage it all. And car companies can still be much more more efficient without fully automating as Tesla learned to its cost.

We already saw that fully self-driving cars are probably much farther into the future and that AI may or may not live up to its hype. We are just too happy to extrapolate from initial successes to science fiction. This is not new. Flying cars, nuclear reactors heating the home, robots having intelligent conversations with us — all of these were predictions built on the assumption that we can safely extrapolate from early successes. We just never know exactly what the next step will bring and when complexity will hit us in the face. We could truly be at the start of a revolution, or we could be just having a very good start to building a ladder to the moon.

My suspicion is that the complexity of most of the problems automation is supposed to resolve is just too high. Over Christmas, I tried to set up a lot of routines to tie together a lot of my smart devices with my daily rituals. It took a lot of time and it was only a partial success — even my simple needs are too complicated for Google Assistant without a lot of manual set up by me. IFTTT, Zapier and Microsoft Flow (now PowerAutomate) have introduced a lot of automation into our online lives but the complexity of the processes they handle can increase exponentially. I’ve used all these services and they make many processes easier. Everytime I highlight a bit of text in an article I’m reading in Pocket, it gets saved to Evernote. But the success is always limited by my ability to conceive of the processes to be automated and integrate them into my life.

The fear is that very soon almost all jobs will be automated away. That is almost certainly not going to happen over the next 10 years. There will be areas where wonderful advances will reshape parts of our lives, but I predict that their overall shape will still be recognisable and not all that different from 2010.

Smart spaces

Much of the technology we take for granted only exists thanks to a quiet revolution that started in the 2000s. It is the revolution in sensor design: gyroscopes, proximity sensors, accelerometers, ambient light sensors, and more. These sensors are cheap, small and can send digital signals to a processor. They are what made the iPhone possible and much of the technology we take for granted would not be possible without them. The big question is what is the next step?

Smart devices, smart rooms, smart homes and smart cities — those are all things people dream about. They dream of cars that talk to each other and the road that tells them how to turn. Cars that can travel closer to each other at higher speeds because they can communicate with each other. Or shops without any assistants. You just walk in, take what you need and the shop knows what you took, will charge your card and order new stock. And how about a home that knows you’re near, will start the heating, turn on the lights depending on the time of day, open its door when it sees you, and so on. And why not a smart city — while we’re at it? A city that knows where there’s trash on the ground, where to turn the lights on, where there’s congestion, where to send the cleaning crew and where to send the police.

Some components of this future already exist. There are offices that know whether a room that was booked was also used. Amazon and others are experimenting with shops that don’t require any staff. Smart thermostats and smart blinds are also something people can have in their home. Smart cities are being experimented with. So, what can we expect in 2030? Probably a lot less than we imagine. Some new things will become common-place — maybe it will be smart locks or drone deliveries. Maybe it will be something not yet on the horizon. But it will be slower and less all encompassing than we imagine.

One thing that the 2010s brought into our lives is the self-service check-out till in supermarkets. That is a change some people love and others hate but it is instructive in what can be done. First, self-checkouts are only possible because of a sophisticated technological infrastructure — the shops does not only have to have accurate stock but also have more information in the database about product weight, special offers, etc. The self-checkout machines themselves have to have very accurate sensors for weight and fairly reliable barcode readers. But the self-check out mostly depends on people learning how to do simple tasks guided by an interface. It is as much a social framing change as it is a technological change. And when it comes to smart spaces, they will always have to have a social component — and only those changes that can be accommodated by people will be the ones that will be part of the overall shape our lives in 2030.

Technical debt ebbs and flows: IPv6, ARM transition and more

“Nothing hides problems like success” is my favorite phrase I learned by listening to the Hypercritical podcast with John Siracusa. Apple, Microsoft, Google and many others are hugely successful — their products are popular, sell well and their stock is riding high. This makes them reluctant to tinker with core aspects of their operations. Why fix something that is not broken, or worse, why risk accidentally breaking something that is essential to that success.

This is how companies accumulate what is called technical debt. Slowly, all the little shortcuts and hacks done to make something work add up so that making a small change becomes almost impossible without having to rebuild the entire system. Every successful technological company has a mountain of technical debt and they constantly have to juggle it around. But the same goes for entire countries whose existing infrastructure is tied to a certain way of doing things — and making small sensible changes becomes very expensive and time-consuming.

The US was a pioneer in the use of credit cards and the entire country’s infrastructure is built around payments with cards. But it was built around the old less secure and now less convenient magnetic strip cards. The transition to chip and pin and then contactless payments is taking much longer than in other countries precisely because the technical debt was too high. The reason for this is that the system grew organically and has too many interconnected parts that cannot be blocked off during the transition. At the same time, it performs its central function reasonably well, so the motivation to move to a better system is less than if it were irreparably broken.

You can see another manifestation of technical debt every time you try to open an advanced control panel in Microsoft Windows 10. In 2012, Microsoft released Windows 8 which introduced a new approach for managing the operating system settings. It had a new visual esthetic, it was in a more unified interface and focused on important settings first. The new control panels were a usability improvement. But Microsoft only had time to change the most important settings and it took only about a click to get to the old interface dating back to Windows XP (or even 95). Eight years later in 2020, the process still isn’t finished and it’s not clear it ever will be. There are just too many corners of the operating system — some of them can only be called by using the command line — to get to them all. And this is a relatively mild case of debt because it does not stop Microsoft from going further in a fundamental way — but it gets in the way. The company is right now experiencing a similar issue with moving to a new (and much better) approach to Sharepoint (not something most people will experience).

There are countless examples of technical debt all around us. Much of the innovation in the next decade will therefore not only have to deal with solving specific problems, it will also have to deal with the technical debt of the world around it.

Some of this technical debt goes back to the foundations of the internet. For example, the world has pretty much run out of IPv4 addresses (unique number that must be assigned to any device connected to the internet). There were only about 4.3 billion addresses available in the first place — which is obviously not enough because there an order of magnitude more devices in the world. The only reason the Internet still works is something called NAT (Net Address Translation) — which puts a bunch of devices behind a device that sends out requests to the internet and then passes them on. That’s what your home or corporate internet router or modem do. They also have the advantage as acting as essentially a firewall which is why it is a lot safer to connect to the internet this way than directly. But there are now not enough addresses even for NAT.

Everybody knew about this problem back in the 1990s. Which why IPv6 was introduced in 1998 with a practically infinite number of addresses — enough to put every grain of sand on Earth online. But even by then IPv4 was so widespread and baked into hardware that was powering essential infrastructure and the switch would have required coordination on a global scale and cost trillions. So instead, the world is slowly transitioning (through a lot of effort and coordination) — but even today the majority of internet traffic (and all on home wifi) is over IPv4 which was introduced in 1983. Over the next decade the transition will continue. It will not be in the headlines or even noticeable to users. But it will be a subtle drag on other innovation as important resources are diverted and some of the smartest engineers on the planet are occupied with this rather than something else.

Much of the next decade will also be dealing with technical debt connected to security. The Internet was designed without any security in mind. DNS (Domain Name System) — the system that translates IP addresses into human-readable domain names — is completely insecure and easily subverted. Without it typing www.google.com in the browser address bar would do nothing — the web would cease to exist. So many slow upgrades and changes to it are being proposed and much of the next decade will focus on dealing with this.

Another security legacy is the system of usernames and passwords. It is clunky and a major source of breaches at all levels. We’ve already seen a great growth in ‘second factor authentication’ over the last decade — something the bank does everytime it sends you a code in an SMS message (which has its own problems). The push to replace or augment passwords will continue over the next 10 years. It will be slow and incomplete but I predict that by 2030, we will have better, more secure ways of logging in. I’d like SQRL to succeed but most likely, we will be faced with a portfolio of authentication methods that will still include passwords in one way or another.

One more technical area that will probably define the next decade is the transition to the ARM architecture in the processors that power computers. ARM-based processors are more energy efficient and have other advantages over Intel’s architecture. The smartphone and IoT world started out with ARM and showed the way. But all desktops and laptops run on the Intel x86 architecture. This would not be a problem if we didn’t want our devices to have longer battery life. The 2010s were filled with rumors of Windows on ARM and the first attempt was a disaster. The problem is that software written for one type of processor don’t work on the other without some (sometimes significant) reengineering. But the latest Windows on ARM devices are much better and the rumors that Apple is switching to ARM are growing stronger. So it is inevitable that much of the 2020s will be consumed with efforts in this direction. Unlike, IPv6, this will have a direct effect on the kind of devices users can buy but it will also require a lot of effort to be spent on dealing with the technical debt accumulated over the last 20 years of software development.

Socialising change and paying down the UX debt

Like technical debt, the much less-well known UX (user experience) debt, is usually applied to individual products or companies. But an entire industry can also accumulate UX debt. It comes in the form of typical user expectations and socialised skills. To learn to use an interface, users have invested a lot of learning effort. This is not just personal learning but also learning from others and learning who in their sphere of friends and acquaintances can help them. So often even interfaces and methods that are more efficient to use end up being slow to catch on. And this gets in the way of innovation in product design and makes revolutionary changes rare.

Some interfaces like windows and files in folders are so entrenched that moving to better ones is a slow process. Designers (inspired by psychology) often talk about affordances — how the objects or interfaces present themselves to us for interaction. But digital objects don’t really have all that many ‘natural’ affordances in the same way that a mug has a handle. So, they need to be learned and when it comes to large-scale change socialised.

Often, when we’re at the crest of the innovation wave in some area, everybody who’s using a new technology and is excited about it, has already internalised all the affordances. And it is hard for them to imagine how they are not naturally obvious to everybody. For example, what could be more natural than interacting with objects on the screen with touch? Yet, there’s nothing intrinsically obvious about the pinch-to-zoom gesture popularised by the iPhone. That’s why much of Apple’s advertising featured people pinching and zooming.

When Microsoft rolled out Windows 8, it completely redesigned all the key affordances with a focus on touch. It introduced a really useful and clever swipe from edge feature (if you had a touch screen). But it did nothing to socialise the feature, which gave us videos, like this, of people who would sit in front of a computer and not be able to do anything with it until somebody explained to them. Windows 8 was a usability disaster not because it was hard to use but because it required relearning. And this was not supported by efforts to socialise the new way of doing things. Apple recently replaced the hardware button (an affordance so completely socialised that it feels natural) with a swipe up gesture. Exactly the same change — but supported by massive advertising efforts, conversations, as well as useful on-boarding of users. And it was successful.

Sometimes a change is so easily learnable that any user can figure it out on its own. I’ve noticed a recent change in calendars used by online booking services. It used to be that this required 2 interactions (and it is still common) — but the best have now converted to first click being the start date and second click end or return date. Nothing in the interface signals that but the outcome makes it clear what happens and no user is confused. But such a change is rare. Most of the time, change takes a long time to become ‘natural’ and to change the shapes of our inner scripts and scenarios.

Whatever the next decade brings to change our lives, it will include a lot of learning, relearning, conversation and bafflement. All of the big transformations I listed for the 2010s are trailing behind them a thick long tail of people who still haven’t been socialised to them. People, who save their files only on a USB stick without an online backup, people who still don’t like touch. People who haven’t figured out how to incorporate audiobooks or podcasts into their lives. People who still print out everything they read. And so on. That is to be expected and not all of it is bad. But this is our social user-experience debt. The 2020s will definitely bring new things but much of the effort will have to be spent on socialising the changes brought by the 2010s.

In praise of a boring future

It is the new and unusual that generates news headlines. But I’m more excited about the boring and every-day change that is cheap, easy to use and available to as many people as possible. That’s why I am most looking forward to innovations in product design, manufacturing and commoditization. I remember spending a lot of time agonizing over the specs of my computer, now I know that anything at a certain price will fit my needs. Any camera, smartphone, office software will work out of the box and match my core needs. Sure, productivity experts will still pore over spec sheets and make many painful compromises about the gear they buy. But that is fine, that’s how progress is made. Typical users will not have to care.

On the surface, trains, planes and automobiles look very much the same and perform their core functions in the same way they did 50 years ago. Because they are more subject to fads and fashions, cars have changed the most in their outward appearance, but their speed and mode of operation has not altered much. If you sample the Computer Chronicles (now available on YouTube) between 1982 and 1997, you will see dramatic changes in the way computers looked and work. But relatively little has changed since then in the overall shape of things that was not incremental. Phones underwent a similar transformation in the 15 years roughly between 1997 till 2012. But now the changes are much harder to perceive at any remove. Social networking was in great flux between 2000 and 2015 but nothing much new will surprise us in this field in the future. There are only so many shapes to be explored and while there will be change, it won’t be as radical. And so on.

I love all the little improvements and changes tech makes in my life and try my best to help other people take advantage of them. But it’s important to be realistic and wait for things to settle before declaring that everything is New Again. Luckily, it almost never is.