Our Current Predicament

The more I study technology, the more difficult it becomes to give a short answer to "what do you think of technology?" or "what do you think of technology X?". So I put together this page to collect some of my central ideas around these questions.

I consider myself a technology critic. When people hear that, they tend to think I'm against technology, but the truth is more nuanced than that. After all, movie critics are not against movies. If anything, I'm against the way we conduct business around technology. So maybe a better term to describe myself would be "techno-ecologist" - one who studies the relationships of organisms to their technological surroundings.

My studies have lead me to some interesting observations. For instance, that the reaction between society and technology is a chemical one - once a tech enters society, it cannot be taken out, only changed and hopefully improved upon. In other words, we can shape the trajectory of technological development by building better and more sustainable alternatives.

They have also lead me to some important questions. For example - knowing that convenience is the main driver in technology adoption - are there certain things that we would still put personal effort into and if so, what are they? In other words - if technology allowed us to automate everything - would there be things that we would rather not? And what would our choices say about being human?

Can technical systems teach us something about how we should live our lives in the physical world? Are there some novel skills that we should acquire to better thrive in an increasingly technological society? Can machines teach us how to be better humans? What guiding principles should we apply when choosing our technical solutions? Do we as a society even have any say in the direction of technological progress and if not, did we ever?

If technological innovation is the primary driver of economic growth and economic affluence is supposedly good for us, then why is a growing number of us so profoundly unhappy?

Is technological progress truly open-ended or is there a threshold of "good enough technology" - a term we normally associate with developing nations - that technologically advanced societies should strive for and after which we start to witness diminishing returns? What is the true cost of a given technological upgrade across it's lifespan?

Do we really want to build societies and grow individuals that cannot function without a computer and high-speed Internet access? What do the biases apparent in Large Language Models tell us about language and humanity? How does a fragmented, highly personalized one-to-one media landscape affect culture, public opinion and cohesion within a society?

Here are some of my opinions on specific types of technology.

Smartphones

I got into mobile programming in the early 2000s with a particular interest in web services and media delivery. I vividly remember thinking how cool it would be to have mobile devices with more computing power and ubiquitous high-speed internet. Fast-forward two decades plus and that's exactly where we are. And everything sucks.

Don't get me wrong, having little supercomputers in our pockets is an enormous achievement and can be extremely useful, but just look at the effect they have had on society. Do we really want to live in a world where practically every function requires a smartphone? Where people get withdrawal symptoms without their phone? Who decided that this was the direction that humanity must follow?

The smartphone may technically be a computer, but it's a computer with very different design goals. A computer is a modeling tool, an active, empowering device, a creativity multiplier. The smartphone is a consumption device primarily engineered for passive observation and entertainment. If personal computers were a step forward from television, smartphones are very much a step backward. In fact, I would argue they are little more than TVs with (crappy) keyboards.

Personal computers were the first truly democratic means of production of the industrial age. A promise of freedom, personal agency and mass productivity. The smartphone is none of those things.

Social Media

I truly regret and despise what social media - and for a large swathe of the population - the entire internet - has become. A surveillance machine, funded by the most toxic business model. An uncontrolled psychological experiment unleashed on an unsuspecting public. An oligopoly stifling competition. They started our race to the bottom and so here we all are.

The last time I logged onto Facebook was in 2016. The last social media website (if you can call it that) that I used was Github, which I quit around 2021. The reason was simple - I realized I was already on the best social media platform of them all - the open web!

I would love to tell everyone to get "off their socials" and follow my example, but the truth is sadly much more complicated. While we are indeed free to change our behavior and choose whatever technical tools we want to engage with in our own lives, the rest of the world will keep moving "ahead". You are going to miss out on things, but I like to think that all the truly important things in life will still happen. Think of it this way - if you get invited to a party - do you want to be there because your presence was important to the other person or simply convenient?

Social media should be about untold numbers of different software systems, each tailored to the needs of the specific customers all working together exchanging information. It should be about us maintaining full control over our data and us deciding the terms of service.

The fact that a handful of service providers can serve billions of simultaneous users is extremely impressive, but that fact doesn't really make a platform more useful. Unless of course you're using it for something other than personal communication, which is a whole other topic onto itself...

Networks connect people, platforms separate us. A truly social media platform would be an inclusive place where you can communicate with people regardless of whether they are "on the platform". The older network services were primarily designed like this - email, the web, Usenet, IRC, etc - all open standards supporting interoperability. For years, proprietary systems tried unsuccessfully to replace them, until the smartphone revolution provided the right opportunity.

Digital Infrastructure

There was a time when people built and maintained servers and networks. I miss those times. Now everything has to be abstracted and automated. Regardless of whether all the added complexity and cost is actually worth it. If Google's doing it then your shop should be too! Why bother setting up your own email server? DNS is hard! Backups are tedious! Yes they are, but they're also important. That's why you take personal responsibility for them. You make mistakes but then you figure things out.

This is important because when we don't run our own infrastructure, we also don't have full agency to innovate. We trade freedom for convenience. We also set society on an unsustainable path of dependency on third parties for even the most trivial of tasks. System Administrators will no longer have full access to their systems and will be reduced to Google Workspace (or whatever other product) operators instead. Trading knowledge of a technology for knowledge of a product. You can transfer knowledge of a tech to a new environment, not so much for a product.

Just imagine how much more prosperous society would be if each country had thousands of email providers instead of everyone using Google Mail.

When we outsource all of our digital infrastructure we also lose our objective view of reality. We don't see what's actually happening on the network, only what our service provider wants us to see.

When you don't control your infrastructure then you don't really own your digital identity. Your Instagram @username is not really yours - you cannot move it somewhere else. It only works because Meta makes it work and allows you to use it. Same thing with your email address - you can't take your gmail.com address "with you" because you don't own gmail.com - Google does. We used to have the same problem with phone numbers until legislation in the 2000s made it possible to switch phone operators without changing the phone number. Maybe some day we could see that happen with other communications networks as well.

We see the same problems with digital literacy. For example, knowing how to write HTML is an incredibly empowering skill (not to mention knowing a "real" programming language), yet I don't know practically anyone outside the software industry that knows how to do it. And the reason is - again - money. There's simply much more of it to be made renting tools and platforms. To use the old analogy of the fisherman - knowing HTML allows you to not only learn how to fish, but also how to build a fishing rod while using Facebook is like getting your fish in exchange for all of your personal data.

Instant Messaging

IM as a concept is great. I tried really hard to use email instead, but it's just not a good fit for today's near-real-time modes of communication. We knew this decades ago. IM is a great way to share data with people you already know.

My beef with the current situation is however twofold - that our current IM platforms are built on unsustainable business models and that there is no interoperability between them. WhatsApp, Telegram, Signal, iMessage, etc - they all do essentially the same thing, yet a customer of one cannot talk to a customer of another. Why is that?

Also, the apps just plain suck. We've had answering machines since the fifties, but I cannot set automatic reply in my IM app in 2025? Or have messages automatically forwarded to a different service?

Another problem with our current situation is that it stifles investment into open standards. The only open, cross-platform instant messaging protocol that's supported by all telcos - SMS - is woefully outdated and hasn't served the real needs of the people in nearly two decades while mobile data technology has evolved several generations over that same time period. The telcos' greedy policies of charging for individual messages and roaming charges are partly to blame, driving people towards IP-based solutions, but the biggest reason is the fact that all the progress is happening in those IP-based apps. It's essentially the same vicious cycle plaguing email - it sucks, but there's no money in fixing it, so all the investment goes into trying to reinvent it in some other form.

Machine Intelligence

I really dislike the term "artificial intelligence" because it implies some sort of reasoning capability or likeness to human intelligence which doesn't actually exist. Illusionary Intelligence (II) would be a much more fitting moniker. ICT exhibits this rather disturbing pattern of releasing unfinished products, making bold promises, disrupting established sectors and norms and then patching the product until it maybe even works (albeit very differently from the initial plan). Then, when it's time to release the next version, throw everything out and start over.

MI systems follow the same pattern, just with much more devastating consequences in these early stages. The pot of gold on offer to whoever plants their flag first on this new territory is just so lucrative that any adverse effects on society, education, mental health, employment, energy or water usage becomes completely irrelevant.

The biggest effect and the one that's so far had the most direct influence on my own life has been in education. MI adoption has been so rapid and widespread, that I might now be the only person in my class to not use any MI tools. I believe the primary goal of school is the work that you put into your schoolwork, not the diploma you get at the end - that's secondary. I write to communicate ideas and to become a better writer. I cannot do that if I let something else do the writing for me. Also, why would I expect someone to put the effort into reading something that I didn't feel putting the effort into writing?

The sad thing is that, as a technology and research tool, language models are very interesting. I think they can really help us further our understanding of some complex problems. But speed-running through your degree is not one of them. There's a lot more that I have to say on this subject, but I'll come back to those at a later date.

The Technological Challenge

What do we mean when we say things like "AI presents a challenge to our education system" or "automation is challenging our manufacturing sector"?

Typically discussions that follow such questions deal with identifying and minimizing the negative consequences of said tech. But that's not really what "challenge" means. It is not a synonym for "problem":

challenge / noun / a call to someone to participate in a competitive situation or fight to decide who is superior in terms of ability or strength.

It's actually a positive thing! A call to action for us to do better. Not to figure out how to adopt something as quickly as possible or how to retrain our workforce. It's primarily about improving our ends, not just iterating on the means. So when "AI challenges our education system", we should seize the opportunity and take a few steps back to think hard about our definition of a "good education system".

In tech, we spend so much time obsessing about the "how" that we completely forget the much more important question - "why". I believe the first step in tackling the technological challenge lies in the answer to that question.

Our Lost Values

For millennia, religion served as our main source of moral and ethical guidance. The scientific revolution of the late 17'th century reversed that trend as we started to look towards science and technology for solutions to life's greatest questions and challenges.

One of the fundamental issues with our relationship to tech is that it lacks moral guidelines. Tech is morally ambivalent and development is driven primarily by corporate entities which are amoral by nature. People on the other hand are highly moral creatures - notions of right and wrong are the foundation of not just interpersonal relationships, but of humanity and civilization as a whole. This creates a big gap in our existence - between what we consider right or wrong and what technology enables us to accomplish. Medicine gave us much more in terms of reducing human suffering than prayer ever did, but by abandoning the gods, we also abandoned an important moral authority.

As a devout atheist, far be it for me to advocate for more religion or some sort of return to "traditional values". Ethics and morality are much older and more fundamental concepts than any religion and it is possible to live a moral life without the church. Ethics and morality created religion, not the other way around. All we have to do is set our priorities straight.

With that in mind, I would like to propose some values to base our technological society on, namely: diversity, resilience, simplicity, modularity, efficiency and accountability.

Diversity

I would like to see much more diversity in our technological ecosystem. Technology is - among many things - a grand competition of ideas, but one that's sadly been dominated by the same handful of product, service and business model paradigms. Which is not to say that alternative ideas aren't out there - they just tend to get drowned out by the mainstream.

We should have thousands of different communication platforms to choose from, not just a handful. Diversity promotes experimentation, interoperability and standardization. Nature abhors monocultures and tech should do the same. Centralization of power tends to lead to stagnation and corruption and technology is not immune to this tendency.

Diversity of ideas is the essential prerequisite for all the following values which is why I believe technology policies should primarily focus on fostering diversity. In addition to holding Big Tech's feet to the fire, we should be doing all that we can to promote better ideas and help them thrive. Letting the private sector handle most of the funding tends to skew the priorities and select for disruption and ultimately for monopolies.

Resilience

Our quality control processes for software development are woefully inadequate. We should devote much more effort on making existing tech work better instead of constantly trying to push for something new. Just throwing more and more management and monitoring systems on top of an already unsustainable heap of complexity just makes things more unstable and insecure.

Being resilient also means being less dependent on technology.

Simplicity

Tech should be as simple as possible. This applies not just to the user interface, but the entire architecture of the system. "As simple as possible" is difficult to achieve mainly because to know what to leave or take out of the system requires much more insight and experience than adding things into it.

Our main tool for achieving simplicity is of course design. Design in the broadest sense possible - the modeling of something.

Resilience and simplicity work hand in hand.

Modularity

One of the reasons we keep seeing growing complexity in our technology is because we cannot recombine what we already have in useful ways. So instead of combining A and B to achieve C, we resort to writing C from scratch or bolting it onto A or B.

The big reason we spend so much time on the error-prone process of maintaining and extending existing systems is because those systems were never designed to be modular and extensible. We have come a long way in modularity on the code level, but this is more about modularity on the system level. I would even argue that modularity on the code level results in several problems with simplicity and resilience because it incentivizes us to write more new code.

For a good example and an excellent blueprint for how to sustainably build large, complicated systems from modular parts, look no further than the *NIX architecture.

To put it another way - unsalted butter is more practical because you can always add as much salt as you need later.

Efficiency

The technology we use should use as little resources as possible. This goes doubly so for critical services. Yes, developer time is also a valuable resource, but natural resources are more important because they are actually finite. I would like to see us move away from running what are essentially rapid prototyping tools (stacks of highly dynamic interpreted languages) being used in production on a massive scale. Moore's Law was an anomaly and an unprecedented privilege, not a law of Nature. A newer version of a system should always use less resources than the system it supersedes, not more as it typically does today.

Case in point - we should replace most of our WordPress installations with static sites.

While we should increase our efficiency in our systems, we should decrease it in our organizations. Designing efficient systems requires creativity, but creativity is at odds with efficiency. Creativity may thrive in scarcity, but scarcity and efficiency - while related - are still two completely different things.

Accountability

Tech vendors should be held accountable for the stuff they release and push onto society. The current uncontrolled experiment on the public simply has got to stop. Commercial software should come with a warranty.

Engineering schools should adopt the Archimedean oath as part of their graduation process.