It’s my birthday. Improbably, I’m 44 years old. That’s what the calendar says. I certainly don’t feel 44 years old. But at the same time, the facts point to yes: I was born in the seventies. I can remember the Challenger disaster really clearly. My first computer was a ZX81. When Congress kickstarted the commercial internet by passing the Scientific and Advanced Technology Act - and the term “surfing the net” was coined - I was thirteen. When Myspace was launched I was twenty-four. Instagram launched when I was in my thirties. (But aren’t I still twenty-four? Aren’t I in my thirties right now? Didn’t everything that’s ever happened to me just happen today?)
With all due respect to people who are older than me, and in particular to that one guy whose mind I blew when I was on high school work experience because I was born after Star Wars:
What the actual.
Anyway, I’m forty-four, apparently.
Somewhat arbitrarily, but in the interests of looking forward instead of back and of marking my birthday here in some way, here are 44 thoughts about the future. You might agree or disagree. Take ‘em or leave ‘em.
One. I think about the climate crisis a lot. It mostly seems overwhelmingly daunting. It’s one of those things where it’s hard to know if anything you’re doing is making a difference at all, and you kind of have to trust that it is because the feedback loop is measured in decades, and it requires so many people to pull together. You have to be optimistic because if you don’t you’ll just give up; you have to keep doing stuff because it all matters.
I think we’ll pull ourselves away from the very worst of the climate crisis but I do think it’ll get much worse before it gets better. I think the definition of “we’ll” here is important: it won’t be my generation that does it. And we won’t save ourselves by trying to financialize our survival. Markets will bring us closer to extinction, not save us from it.
Two. I think the (re-)rise of authoritarianism and the increasing importance of the climate crisis are linked. It’s not an accident that Bolsonaro was in favor of felling the Amazon or that Trump had such a strong fossil fuels agenda. If a wealthy industry feels like it might be politically under threat, it’s going to do everything it can to change the politics and create a context where it is protected. In a way, it’s a version of the same international politics that have played out for hundreds of years: proxies for the interests of wealthy individuals who don’t care about anyone’s well-being but themselves.
We’re going to need to break up those companies.
Three. I don’t buy into the metaverse’s Ready Player One vision of a three-dimensional virtual world we all inhabit. I think it’s silly. But behind all that, there’s an interesting interoperability story. In fact, there are two dueling visions:
The first is Meta’s. It wants to own the operating system for online experiences. You’ll be able to take artifacts from one experience to the other, and it’ll all be powered by Meta’s underlying technology.
The second involves NFTs. You’ll, again, be able to take artifacts from one experience to the other, powered by your wallet and a standardized contract for metaverse NFTs. No single company’s work powers it, but it does rely on blockchain and contract standardization.
Interoperability has been an issue since the birth of the application web. Although desktop software has interoperated for decades using file formats (consider the number of apps that can open a text file, say), web apps tend to be data silos. I think it’s positive that multiple companies are putting money into building online interoperability: although they’re all likely to fail at what they’re setting out to do (I really don’t think either Meta or NFTs are the solutions here), something good may come of it that others will be able to use.
Four. It is unlikely that AI solutions like ChatGPT will be a major part of the future for a few reasons. Firstly, ChatGPT is a specific type of AI language model known as a transformer model, which is designed to generate human-like text. While these types of models have achieved impressive results in generating realistic-sounding text, they are not able to truly understand the meaning of the words they are generating. They are simply following patterns and rules that they have learned from large amounts of text data, and they do not have the ability to comprehend the context or meaning of the text in the same way that a human does.
Additionally, AI language models like ChatGPT are limited by the data they are trained on. If the model is not trained on a diverse enough dataset, it may generate biased or inaccurate text. This can be a problem for applications that rely on the accuracy of the generated text, such as in the case of chatbots used for customer service or information gathering. In these cases, it may be more reliable to use a human operator rather than relying on an AI model to provide accurate and unbiased information. While these models can be impressive in their ability to generate human-like text, they are not a substitute for human intelligence and should be used with caution in applications where accuracy and fairness are important.
(This entry was written by ChatGPT.)
Five. It seems inevitable that broadcast television will be taken out of commission and replaced with the internet. Not immediately, but in the next 15 years or so. The thing is, TV is essentially free: you buy a set and you can receive content with no further outlay. On the other hand, the internet is an ongoing monthly expense - and a pretty significant one in countries like the US. Removing free-to-access over-the-air content will create an information void.
Absent legislation that prevents it, some companies will fill that void with free-to-access internet connections that prioritize their own content, a bit like the one Facebook tried to create in India before it was run out of town. Should it happen, those companies will essentially end up owning media consumption for a large chunk of the population.
Six. I think we need to call it: email’s not going away, as much as we wish it would.
Seven. For decades now, there’s been a meme in education where folks want learners to put together a portfolio of their work to show employers. They’re often called e-portfolios because the idea dates back to when people felt you had to prefix internet stuff with an “e-“.
I’ve done enough hiring now to wish we had something a little different for engineers. I would love to know that a candidate has created a product from scratch: they’ve done some research, tested an idea, created a prototype and tested that, and then worked to build it. That kind of prototype-driven product mindset is hard to find in engineering, and giving candidates the space, time, and tools to do this would set them apart.
An e-portfolio where the candidate demonstrates how they think would also be lovely. It’s called a blog. I always find it really helpful when candidates write their thoughts in public. In a world dominated by neural networks and repetitive work, unique thinking and creativity will set someone apart. Can we encourage public blogging for students again? And can we build product thinking more deeply into the engineering discipline?
Eight. Our communications have become more and more digital, but the cables and signals that transmit them (of course) remain analog. Digital signals are 1s and 0s, but those translate to tolerances in the analog world: a signal within a set of tolerances is a 1, and a signal within a set of tolerances is a 0. I wonder what we can do within those tolerances? Could you hide messages within the transmissions, like a kind of broadcast steganography? What if it was happening right now - in cellphone networks, say? If everyone is focused on digital, which they will be more and more, analog becomes a kind of wild west playground.
Nine. I worry about how to help my baby be a creator rather than a consumer. If anything, apps have put consumer culture on steroids since I was a kid. Will he want to write stories and draw and imagine, or will he just want to watch what someone else has made?
Ten. I’ve always thought there was something really magical about ARGs: games with puzzles that you solve by being out in the world as well as building community online. In some ways, games like Niantic’s Ingress and Pokemon GO have harnessed the core appeal of that idea, but in other ways not. There’s a Sherlock Holmes / Glass Onion appeal to working together with people all over the world to solve one puzzle.
I know people have been writing for decades about how to “gamify” real-world problems, but I wonder if there’s a way to harness these game mechanics for the purposes of global collaboration. For example, if a community in X country needs a renewable energy resource so that it can do Y, what if there was a way to get people together to help them build and maintain that, according to peoples’ abilities all over the world? Outside of any financial compensation or other financialization?
I guess what I’m asking is: what if working together to help people get through the climate crisis could be fun? Does the apocalypse have to be dour? Would more people take part if there was a spirit of hopefulness and community? And if there was a sense of pride, achievement, and maybe even acknowledgment for helping people in need?
Eleven. In a world based on profiling, probabilistic prediction models, corpus-based decision-making, and near-ubiquitous surveillance, only people who don’t conform to the models anticipated by the people who built and designed the systems and therefore aren’t tracked as closely can really be free.
Twelve. Atoms have value. Bits, relatively speaking, do not.
The value of digital information is a function of the change it can effect in the real world.
A digital book is valuable for its effect on you. A physical book is valuable for that, but also in itself: for the binding, the typography, perhaps the illustrations, the quality of the printing, and its feel in your hand. A well-made physical object will always be more valuable than its digital counterpart.
For most objects, considering value in terms of scarcity is missing the point.
Or to put it another way: I’m short on NFTs and long on hardback books.
Thirteen. If linear time is an illusion, what if the illusion itself is developmentally achieved rather than inherent? What if newborns can see the entire vastness of time but have no way to comprehend or convey it? Omniscience is wasted on the young.
Fourteen. Public transit is one of those things that I think should be the measure of any society. It’s a kind of equalizing infrastructure that you’d need to be incredibly dogmatic not to put into place. For example, the reason a lot of the US has poor public transit - and why some people are anti-bus and anti-train - is because of real spending by the automotive industry rather than any practical reason.
At the same time, it has limitations: you need to live close to a stop, and it has the potential to leave out people with real mobility problems.
Assuming self-driving vehicles become a reality on every public road, I wonder if swarms of vehicles could be the future. You would summon an autonomous pod on demand that could then join a “train” of other pods that could share resources as they headed along trunk routes. On longer journeys, there could be specialized dining pods, sleeper pods, and so on. And then you’d head back to an individual pod for the last mile to your destination.
The pods could be provided through a public-private partnership, as buses are right now in many cities. It wouldn’t work as a fully-private endeavor: even if pods were provided by multiple companies, they’d need to be compatible, and localities would need to collaborate on defining the trunk routes. Private businesses could provide different eating and sleeping pods, for example, but they would still need to adhere to the rules.
Fifteen. There will be apps that translate the cries of a baby, the meows of a cat, etc, into meaningful notifications. “Your baby is bored”; “your cat wants love”. Rather than helping to build deeper bonds, they will create more emotional distance by getting in the way of our empathetic intuition.
Sixteen. Jaywalking will eventually become legalized everywhere. As it should always have been. This will be an unrolling of a century-old automotive-sponsored campaign that transformed American cities in ways that were against the interests of their inhabitants (but very good for cars).
Seventeen. I keep coming back to Pascal Finette’s talks about exponential thinking. Change is constant but, taken as compounded impact, hard to imagine.
So much of our understanding of the world has been shaped by commercial interests, whether it’s an imperial desire for more resources or an industry seeding the idea that its alternatives aren’t fit for purpose. In the US, a lot of people think that socialized healthcare is unworkable because of what amounts to a long-running PR campaign that isn’t based in fact. The same goes for car culture and gun culture: it’s all marketing.
As our needs change, marketing changes. And every incremental movement towards a safer, more equitable culture matters.
Rural America’s guns-and-trucks culture might seem inevitable now, and we might be (rightly) worried about its accompanying Christian nationalism, but if people keep working for justice, I think it’ll be gone within a generation. Likewise, the private health insurance industry might feel like a part of the furniture, but we’ll look back at it as ancient history by the time I’m a grandparent. This, too, shall pass.
Eighteen. I’m very bullish on more federation, particularly around social media. Arguments - which are abundant - that people are too lazy and that the user experience will never be good enough remind me of the early days of the web. For many consumers, the web replaced CD-ROMs and walled gardens like AOL and CompuServe, which were all initially more coherent, better-designed experiences. But the expansive building-block nature of the web very quickly won out. The possibilities with federation are exponential, as we’re beginning to find with purpose-built platforms like Pixelfed and alternative front-ends like Elk. Just wait.
Nineteen. Given the confluence of a recession and the energetic re-emergence of end-user open-source tech, I think there’s a good chance that the majority of transformative projects on the internet in the next five years will not be venture capital funded. They may well not be businesses or traditional organizations at all. The only way to make money from them will be to spot the trends they create early and meaningfully participate in their ecosystems. But making money won’t be the point at all.
After that five years, we may see VC-funded startups that build on these grassroots projects start to take hold. How they affect their parent ecosystems will depend, at least in part, on the designs and governance of those ecosystems. It’ll be up to project maintainers to protect themselves from being absorbed into a commercial entity - if that’s what they want.
Twenty. There are, very broadly speaking, two dominant kinds of software programming:
Software engineering - where a group of people work on a shared codebase using shared standards and accepted best practices for reducing risk, including continuous testing and deployment. Usually, this is part of a larger team that includes a product manager, among other roles.
Personal expressive programming - where one person writes software for their own purposes, driven at least as much by the creative pursuit as the desire to build a product. There may still be tests, etc, but the result is often an idiosyncratic “personal codebase”.
There’s nothing wrong with either of them! The former has dominated the internet for the last decade or so. As the tech industry re-organizes itself in the face of a recession, I think the latter is about to become much more important. (Expressive programmers might even be able to use AI to later bring their code in line with external standards.)
Twenty-one. I suspect coffee drinking, like meat eating, will become uncool as the climate crisis progresses. If it doesn’t, it should, because coffee production is environmentally disastrous and remains remarkably socially acceptable despite this. Probably because most of us, myself very much included, are addicted. I like both the taste and the ritual, but people used to say that about smoking, too.
Twenty-two. They used to think that babies didn’t feel pain. They thought this until the mid-1980s. I’m going to go out on a limb and say they didn’t really think to ask mothers.
There was also a school of thought that said there couldn’t be sentient thought without language.
Both of these are so obviously bunk from our perch in 2023. As time goes on, I think we’ll learn that full cognition is not as rare as we thought.
But also, what do we hold to be true about cognition and sentience today that will obviously be wrong thirty years from now?
Twenty-three. Are AR and VR needed iterations of the computing experience, or are they needed because someone needs a new paradigm to sell?
I used to be pretty enthusiastic about AR, but the more I think about it, the less I want my reality to be augmented by someone else’s product (and through it, someone else’s design decisions and business priorities). My reality is mine. You can’t have it. You certainly can’t sell it back to me.
That’s not the same thing as being anti-progress. I want new kinds of computers, I want innovation to continue on the internet in particular, and I want to use new and exciting technologies. I just want them to respect my humanity and leave me alone. My thoughts and experience of the world are not an opportunity to sell me something or make me more productive.
Twenty-four. A lot of these entries are about the role of capitalism in the future; it’s certainly on my mind.
It’s not that I don’t think capitalism can have a role. But it’s got to be as part of a balance that includes considering the well-being of all people. It’s not okay to trust the market to protect the lives of the vulnerable, particularly in a world with dwindling resources. Every living human’s right to life must be guaranteed - alongside their civil rights, their right to a home, their right to healthcare, and so on. Any system that allows people to go hungry and prioritizes the profit of a few over the well-being of the many isn’t worth preserving.
This has always been a struggle, but as the planet heats up, we’ll have a real fight on our hands.
It’s odd to me how ingrained the capitalist grind is in American society: people are proud to work until they die. They’re excited to contribute less to their neighbors. They’re islands unto themselves. We’re not going to survive this way.
Twenty-five. I think the reason people are obsessed with using AI to generate art and creative writing is that they wish they could do those things themselves. AI presents a way for them to at least have some control over some artistic output.
What if we taught more people to paint and write and gave them the time and energy to get good at it instead? Or, you know, hire artists?
Twenty-six. I’d love to have six picture frames on my wall that update each morning with the latest photos (maybe from Instagram, Flickr, or their own sites) from my close friends and family. No feeds; no scrolling; just delight. The captions could show up next to them like descriptions in a museum.
I’d love more innovative and ambient displays for web content in general. “Less addictive but more delightful” feels like a good mantra for new interfaces. Less addictive without the delight is boring; more delightful with addictive feedback loops is still psychologically heavy.
I’m particularly interested in stuff that is static but changes when you’re not looking at it: an e-ink display that shows today’s newspaper, for example, or kitchen wallpaper that changes overnight.
Twenty-seven. For AR computing to work it’s going to need to be tactile. An iPhone is such a good personal user experience in part because it’s a piece of metal and glass that sits heavy in your hand; the tap of your finger against the glass creates its own haptic feedback, and then the device has its own haptic functions. It feels real: atoms have value and bits do not.
Calm technology has some examples of haptic AR: a belt that lets you know when you’re walking in the wrong direction, for example. There are new devices for the sight-impaired along similar lines.
Haptics don’t really make sense for AR lenses though: the last thing anyone wants is for their devices to start hitting them in the face. And anyway, we don’t experience the world through sight alone. So AR is going to need to be a whole-body experience.
Twenty-eight. Meat substitutes and electric cars are for people who don’t really want to change in changing times. Disclosure: I drive an electric car and have been known to eat meat substitutes. But electric cars still clog up the roads, and their manufacture and tires still have an environmental impact; meat substitutes are not particularly nutritious.
The situation demands that we actually change our behavior, but that requires more imagination than we’ve seen from mainstream businesses. What are people at the perceived fringes doing? That’s what we’ll all be doing in twenty years.
Twenty-nine. That last statement bears repeating in its own right:
What are people at the perceived fringes doing? That’s what we’ll all be doing in twenty years.
Thirty. I wonder when they’ll announce the first security breach that was wrapped into a neural network corpus? Has it already happened?
“Your personal details were found in a new data breach. Also, they were incorporated into 38 AI engines and used in the composition of 17 poems.”
In all seriousness, if data breaches find their way on the web and AI neural nets use public web data, why wouldn’t your leaked private personal information find its way into a corpus eventually?
Thirty-one. If Hyperloop was designed mostly to draw attention and investment away from shared public transit, as many say it was, it’s worth thinking about which other infeasible or outlandish projects might be used to draw attention away from our collective well-being in the future.
And at the same time, it’s important not to be too cynical: some progress really is progress. Technology can be devised and built that is for our genuine benefit. New ideas can be transformative and liberating. The world should not sit in stasis.
What’s the term for being optimistic about new developments but aware of the potential harms? Optimistic skepticism? Skeptical optimism? Just being awake?
Thirty-two. I’m excited for Automattic to buy Twitter once it’s been fully written down. Matt and his team will do good things with it.
Thirty-three. It’s surprising to me that we haven’t (to the best of my knowledge) seen the development of a new political system or philosophy that incorporates global hyperconnectedness into its core. We’ve certainly seen new political ideas that incorporate startups - the neo-reactionary movement’s concept of a CEO monarch is one, and potentially DAOs are another - but what of just the idea that everyone can collaborate with everyone else? What does that do to representative democracy? What does that do to the concept of nations, even?
There must be existing work here, but it hasn’t been mainstreamed in the way that socialism, say, was a hundred years ago. What does innovation around how society is organized really look like for ordinary people in the current post-industrial context? Can we use connectedness to make the world more equal, or is that just wishful thinking?
Thirty-four. I’m stuck on the idea of helping people to create art.
Kurt Vonnegut:
“Go into the arts. I'm not kidding. The arts are not a way to make a living. They are a very human way of making life more bearable. Practicing an art, no matter how well or badly, is a way to make your soul grow, for heaven's sake. Sing in the shower. Dance to the radio. Tell stories. Write a poem to a friend, even a lousy poem. Do it as well as you possibly can. You will get an enormous reward. You will have created something.”
How can we help more people to do this? I’m serious. Art speaks to our humanity. Making art - particularly for people who have been under the thumb of performative productivity and grind culture - can be a restorative reconnection with being human. All of us need more of this. We need it for ourselves, and we need the people around us to do it, too. Reconnecting to our humanity is how we get to a future that isn’t trying to kill us.
Thirty-five. Rather than proof-of-work or proof-of-stake, imagine a proof-of-effort blockchain. The algorithm would notice when you’d put in a certain amount of manual labor (relating to pre-defined tasks like assembling a product or picking a crop) and would pay you accordingly. Then each worker would no longer be an industrial employee but would effectively be an independent contractor, free to choose their own hours. (The level of pay might dictate that they need to work very long hours in order to make a living wage, however.)
Hopefully, you understand that I’m not seriously suggesting this. But it’s not inconceivable that someone else will. To someone, that sounds like a very good idea. No need to worry about employee benefits or employment law! Blockchain is anonymous and anyone could do it, so no more worries about keeping people safe! It’s like Mechanical Turk but for backbreaking work! There are companies out there that are desperate to get out of the requirement to treat people well.
Thirty-six. Organ transplants are incredibly invasive and often require strong anti-rejection drugs. Even then, they can have a short shelf life, or life expectancy can be low.
There’s some work being done to reduce the chances of rejection and therefore the need for anti-rejection drugs, but invasive surgery remains. The real breakthrough will be when we can regenerate organs in vivo. It won’t help people with catastrophic damage, for example from an accident, but it would change the game for people with long-term degenerative diseases.
As it stands, major organ transplantation sounds good, and it does prolong lives, but it’s a very intense experience for the transplantee that is far from allowing them to live their lives as before. Eventually, this will change for the better.
Thirty-seven. Some startup is going to turn engineered epigenetic inheritance into a product. Much cheaper and easier than editing your future child’s genetics directly.
Thirty-eight. Some forms of user testing are indistinguishable from human psychological testing and will, eventually, be strongly controlled or banned.
The same might go for “fake it ’til you make it” approaches, which should probably have been more strongly restricted after Theranos.
Like an ingredients list on a food product, it will ultimately need to be a requirement that consumers know what they’re getting into when they start using a software service. And should they be tested on, those tests need to operate within the rules that other scientific testing must adhere to.
Thirty-nine. I often wonder how long the internet will survive. The answer is not forever, and it probably has an end date sometime in my lifetime.
We’ve already begun to see national splinternets, although these controlled networks do still peer with each other. It’s possible to use a VPN from behind the Great Firewall in China, for example, and connect with a server in the US. The real splintering will happen when peering breaks down and we lose communication trunks between nations.
Using the internet as a foreign policy tool to further American interests is a good way to hasten this disintegration. I do think the internet is a force for good, and I’d prefer to avoid its destruction if at all possible.
If the internet does die, I wonder what comes next? I’d love to see more durable citizen-run networks that run across rooftops and mountain ranges. Less reliable than a big cable run by a multinational telco, for sure, but more decentralized and potentially less prone to censorship.
Forty. When our children ask us what it was like to live in the Trump era, will we reply that it was terrible, or will it seem good in comparison to that vantage point in history, even despite the presence of Trump and everything he did?
Forty-one. The epochalypse - meaning 03:14:07 UTC on 19 January 2038, when the date becomes too large for UNIX-based systems to store in its existing integer format - isn’t likely to be a big problem in the same way that Y2K wasn’t a big problem.
That’s not to say that it won’t take work: operating systems will need to be upgraded and file formats will need to be changed. I’m sure there will be a ton of noise about it leading up to the moment, and then when nothing happens, people will assume it wasn’t ever a big thing and it was all overblown. The real truth will have been somewhere in the middle, and software developers the world over will pat themselves on their backs.
It’s kind of neat in a way to have these little moments to look forward to.
Forty-two. Someone sometime soon is going to come out of leftfield and absolutely blow our minds and change everything forever. We’ll be left scratching our heads thinking, “where did they come from?” and it’ll be somewhere where nobody was looking, and the kind of person who is still underestimated, and it’ll be awesome.
Forty-three. I’m probably (in some ways hopefully) just about halfway through my life. It’s not as sobering as I thought it would be. While I can’t exactly say that I’ve lived a life free from regrets, I’m so incredibly grateful for the things I’ve experienced and the people I’ve met. That’s been the overarching theme of my life so far: people are incredible, you know?
That’s been what’s kept me going with the work I do: technology itself is only sporadically interesting, but people are fascinating. In the first half of my life, I didn’t always understand that. In the latter half, focusing on openness and humanity feels not just right but unavoidable. People keep everything interesting; you never know what’s beyond the next corner.
Forty-four. The trick is to give people space to surprise you. Including yourself.
Share this post
