WHY WE NEED TO STOP COMPARING A.I. TO PREVIOUS TECH REVOLUTIONS

We’re in trouble. All of us, and it's a lot of trouble. AI has already destroyed our world, we just haven’t realised it yet. I can already see the eye-rolling and hear the howls of derision from the AI cheerleaders and fanboys. But, a few months ago I was one of you. So, what changed? I realised that most commentators are either ignoring or vastly underestimating, the impact of one specific aspect of the AI revolution. That aspect is the rate of change, plain and simple.

No, The Industrial Revolution Is Not A Valid Comparison

Over the last few months, I’ve read countless posts and articles dismissing concerns about the pace of change by drawing comparisons with previous technological innovations and revolutions. However, none have been valid - they are comparing ‘chalk and cheese.’ You see, what many people don’t appreciate is that any technological change has two distinct major components. One is obviously the technology itself, but the other is how it interfaces with our grey matter - our biological technology if you will. Yes, there is the relationship between the technology and our society, but that is a subsequent result of its relationship with our brains.

Before we go any further, let's put things into some context. We’ll start with the Industrial Revolution - a favorite comparison of the AI advocates. It’s a subject I know quite a bit about, having been born and raised in the English Black Country (and, before any woke types get offended, the Black Country is a genuine region of the English West Midlands, covering areas including Dudley, Sandwell, and Walsall. It got its name from the coal dust and assorted pollution that, quite literally, coloured the ground and buildings during the Industrial Revolution). Due to being ‘blessed’ with abundant coal and other natural resources, it was one of the first areas of the world to rapidly industrialise in the late 18th and early 19th centuries. Its central location allowed first canals, then railways to catalyse this industrialisation, until by the end of the Victorian era it was one of the British Empire’s primary industrial regions. By the time I was at primary school in the 1980s/1990s, the region’s industry was a shadow of its former self, but still impressive. With so much local industrial heritage at our doorstep, we went on many visits to see factories, coal mines, steam engines, etc.

So what? Well, the changes that took place with the introduction of the steam engine, canals, and railways did indeed seem like a rapid and disruptive revolution to those at the time. As cottage industries became obsolete and new factories with powered tooling opened, there was a mass migration from the countryside to urban areas. As the efficiency of steam engines and various machines improved, jobs changed and cities grew even bigger and faster. But - and this is a huge ‘but’ - we’re still talking about a rate of change measured in generations (if we assume a generation to be 20-30 years). It was perfectly possible for three generations of a family to have essentially the same jobs and lifestyle, albeit with gradual evolutions.

Likewise, if we look at the growth of the railways in Britain, it was a revolution, but the network still took decades to mature into something that could be considered critical infrastructure. For example, although the first public, locomotive-powered railway opened in Britain in 1825, it wasn’t until the 1860s/1870s that railways really became the arteries of the country - and as late as the 1930s, the canal network was still commercially viable for goods.

I could go on with numerous specific examples from the Industrial Revolution, but the point is that despite it being often referenced as a comparison with the current AI Revolution, it is clearly nothing of the sort - the only real commonality being the principle of technological change.

Mobility And Electrical Revolutions Took Time

Right, so, let’s tackle another one of the comparisons used by the AI cheerleaders; cars. Well, that’s easy to dismiss. There were European manufacturers like Daimler and Panhard that were series-producing cars by the 1890s. In fact, By 1900, the wealthy European auto enthusiast could pick from many models. However, it would still be a decade before Ford famously introduced the Model T, whose assembly-line production method made cars much more affordable and accessible. But, even then, it was still decades before cars really democratised mobility, becoming affordable to the masses - well into the 1930s in the United States, and into the 1950s and even 1960s in Europe. Heck, in much of the rest of the world, cars only became accessible to regular people in the 1980s and 1990s! Again, cars actually took many decades to truly mature into a technology that could be widely accessed.

The same is true when it comes to air travel. Sure, the Wright Brothers first flew at Kitty Hawk in 1903, but even though there were massive advances in aviation technology over the coming decades, especially during World War 1 and World War 2, it wasn’t until the 1960s and 1970s that international air travel became a viable and affordable option for the majority of people - and even then, only the majority in developed rich countries.

The development and proliferation of electricity, telephones, radio, and television took a similar trajectory. It took them all decades to develop into truly useful and, critically, widely available and affordable products and services.

Even Computers And The Internet Were Comparitively Slow Burns

This brings us to the most misused comparisons of all when it comes to the AI Revolution; those of computers and the internet. Let's not forget that the first true computers were under development in the 1940s, and were available commercially by the 1960s. But, it wasn’t until the late 1970s/early 1980s that useful home and small business computers became affordable and accessible. Even then, many would argue, computers didn’t figure much in most people’s lives until the 1990s - and again, that is only in the developed world. Elsewhere it took much longer. And, the internet is the same. Today it is hard to believe that, despite having been technically available for decades in more primitive forms, most people, even in rich developed countries, weren’t using the internet until the late 1990s and early 2000s.

So, all of these previous technological revolutions, while definitely being dramatic and disruptive, took decades from inception to the point where they had materially changed society. Decades. Again, that is the same as saying generations. These changes took generations. Absolutely, they involved a rate of change that was much faster than in the preceding centuries, but they still involved changes at a pace that was manageable for both individuals and institutions.

AI Advancing Too Fast For Grey Matter And Social Structure

Now, let’s look at AI. While it is true that AI, in a broad sense, has been under development for decades, the only starting point that makes sense in terms of comparisons to other technology is from when the first AI-powered products became available to the public. This is the equivalent of the first commercially available steam engines becoming available, the first public locomotive-powered railway opening, the first motorcars going on sale, etc. Realistically, this point is when ChatGPT went public at the end of 2022.

Within six short months, the number of AI programs, models, services, and apps has exploded. While ChatGPT, Bard, Stable Diffusion, and Midjourney are already becoming household names, there are already myriad AI-based programs optimised to perform thousands of tasks readily available to anyone with a computer, smartphone, and internet connection. Furthermore, many are free, and even the paid services are often very cheap. Basically, in half a year, AI has achieved the level of proliferation and accessibility that it took the other technologies mentioned decades to achieve.

And, as someone who has been experimenting with AI daily since January, I can confidently say that the capability of AI has exploded during that time. Really, it has been staggering. Take, for example, the difference between ChatGPT 3.5 and 4 - it is like night and day. To put this in perspective, in a recent interview on the Diary Of A CEO YouTube channel, Mo Gawdat, former chief business officer for Google X and author of the books Solve for Happy and Scary Smart, said that ChatGPT 4 already knows “a thousand times more” than the smartest human on the planet and that by 2045 AI will already be “a billion times smarter” than humans. Clearly, when it comes to AI, we are dealing with compounding phenomena, resulting in exponential capability growth. Unfortunately, this dynamic is difficult for most humans to understand, as evidenced by how many people are caught out by compounding interest on credit card balances!

So, here we arrive at the fundamental problem with AI. It is not the technological capabilities per se. Or, should I say, it is not the technological capability at any given point. Rather it is the speed of change and the fact that it is only ever going to increase at an ever faster rate.

Why The Lack Of Rational Perspective? Blinded Or Scared?

Sounds simple, right? So, why then are so many AI ‘experts’ making such ridiculous comparisons? Could it be that they are so captivated and enthralled by the capacities of AI that they are simply blinded to the wider reality? Or, could it be that many people are just trying desperately to stay relevant?

I admit, back in January, when I first started experimenting with AI tools, I was initially seduced by the sheer creative possibilities at my fingertips. But, within a few months, after seeing the rate of change before my eyes, that feeling shifted to a fear of being left behind. We’ve all seen the posts proclaiming that you don’t have to worry about your job being replaced by AI, but by someone using AI, and all that shizzle. Hell, I even wrote blog posts singing AI’s praises, just a few months ago.

I suspect the dread I’m now feeling would be familiar to many Russians in the early 1920s, or Germans in the mid-1930s. That feeling of having been swept up and carried along by a wave of popular enthusiasm and shiny promises, but then starting to appreciate the enormity of the monster that’s been unleashed on the world.

Of course, I’m not the only one becoming increasingly concerned about the actual, real implications of the AI Revolution. We’ve all seen the open letter, signed by the likes of Elon Musk, calling for a halt to AI development, or the high-profile resignation of the ‘Godfather of AI’ Geoffrey Hinton from Google, so that he could openly voice his concerns. And, as for the previously mentioned Mo Gawdat, well, he’s so concerned he’s urged those considering having children to hold off for a few years!

But, it’s not just the brightest minds in tech who are now getting cold feet about AI. On 11th June, Sky News ran a story headlined ‘Regulate it before we’re all finished,’ which highlighted how AI is mimicking real artists to create fake tracks that are already all but indistinguishable from the real thing. A problem indeed for an industry that relies on intellectual property rights! And, this also highlights the fact that many who stand to be most affected are still failing to grasp one of the most fundamental, and worrying, aspects of AI - that something of inferior intelligence (humans) cannot possibly hope to effectively, and sustainably, regulate something of greater intelligence (AI).

Perfectly, illustrating this inferior intelligence is Canadian singer (and Elon Musk’s ex) Grimes. According to the Sky News report, she isn’t concerned about AI, tweeting "I'll split 50% royalties on any successful AI generated song that uses my voice… Same deal as I would with any artist I collab with. Feel free to use my voice without penalty. I have no label and no legal bindings." Apparently, she also likes the idea of being “fused” with a “machine” and “killing copyright.” The latter point begs the question of where exactly she thinks any royalties she’s going to “split 50%” will come from.

Destroying Society’s Foundations Without Offering An Alternative

Still, even if the potential downfall of the likes of Drake and Grimes doesn’t cause you to lose sleep, it's the freight train that’s heading for the normal job market and education system that probably should.

Again, going back to jobs, we’ve all heard doom and gloom stories about imminent mass lay-offs, especially for ‘white collar’ workers. However, they are usually countered, often quite convincingly (superficially, at least), by AI cheerleaders proclaiming how there’s no need to panic because the masses can just be ‘upskilled’ to enter the sunny uplands of an AI future. Illustrating this was a post I read last week, claiming that AI was brilliant because it was democratising the creative processes that until now had required talent, skill, and substantial investment of time and money to master. The author went as far as saying AI would usher in the ‘end of mediocrity’, giving everyone the power to make amazing content - implying they could do so even in the absence of talent, skill, and a substantial investment of time and money.

So, at first, this seems great, right? I mean, hell, we can all create amazing things - written content, images, music, video, everything! Yay, let’s all pop open the champagne and party! The world has been fixed. The problem is that this fundamentally undermines, well, pretty much everything that glues our society and systems together.

Take, for example, students. Young people embark on a very lengthy and expensive process of learning, exploring, and honing skills - usually with a formal education system as the lynchpin, but also investing countless hours and their own resources. These students only do this precisely to acquire those skills and experiences, and access the required mentorship and facilities, that the previously mentioned author so readily dismissed. After education, a similar process starts as the graduate enters the workforce. They usually start at the bottom and work their way up. The motivation to put in the effort and build a career is derived from the need to earn more money and better their lives (and, the more optimistic would argue, the need to find meaning and purpose). They are able to do this by improving their skills and abilities, thus making themselves more valuable to a business or organisation.

Of course, you’re probably thinking ‘duh’.

So, why then are so many out there choosing to ignore these fundamental truths? Torpedo a battleship below the waterline and it will probably sink. Sure, it might make a really impressive explosion - you could say the explosion is better than any explosion the ship’s guns could create - but that would be a completely inappropriate comparison that only bore superficial relationships (i.e. a big explosion, fire, and smoke). To me, our society is that battleship and AI is that torpedo.

The unavoidable truth is that our entire society only functions because there is a clear, measurable difference in abilities and qualities. This applies to services, products, creative output - everything. It is this inequality that provides opportunities for people to better their position in life, creates an environment in which businesses can compete with each other, and motivates people to get off their arses and do things - to better themselves and their abilities.

Everything is relative. At art school, one of the things that was drilled into us is that there’s no ‘good’ or ‘bad’ art. Everything is subjective and everything is relative. We form our judgments based on benchmarks in our environment and society. This is necessary to give a structure for things to function. Basically, what people consider to be ‘good’ is very much dependent on what they also consider to be ‘bad’ - and the disparity between peoples’ perceptions gives the variety that makes everything work. AI is, as I’ve alluded to before, likely going to create a homogeneity of output (whatever the content) that is poison to this organic balance that has taken society thousands of years to arrive at. So, far from ending mediocrity, it will ensure the spread of mediocrity.

And, yes, I know, you can argue that, ah well, that will create opportunities for those who stand out by not using AI. You know, like there will be a renaissance of traditional arts and crafts and alike because it will be even more different to the masses. At some level, this is probably true. In fact, Mo Gawdat also mentioned this is likely to happen. But, both he and I think this is likely to be so minor that it won’t really affect the bigger picture. A handful of artists, writers, musicians, etc. may carve out successful niches, selling their creations to ultra-wealthy patrons who can afford to be sentimental. But, for the vast majority, AI will destroy opportunities (or, at least, opportunities to earn a living and achieve any sense of happiness and meaning).

We’ve Painted Ourselves Into A Corner

Okay, I promise I’ll wrap things up now. So, I’ll go back to my original premise, that AI has already destroyed the world, but most people just haven’t realised it yet. No, I’m not worried about AI going all SkyNet on us and constructing killer robots to hunt us down, because I think things will fall apart before that can happen. I strongly believe we have basically just painted ourselves into a corner and are now trapped.

The AI Revolution is already faster than any other technological revolution in history. This ensures that no education system or social and governmental institutions will be able to keep pace. Not even remotely. The very structure of schools and colleges means a student’s path to graduation takes years. Writing courses, recruiting staff, and building infrastructure - all take years. And, there’s no point saying “speed it up then” because it would be impossible to match the speed of AI development. There is a hard limit because we are dealing with still forming human minds. This takes time. It is biological - hence, we have to give time for children and young people to grow and mature, and can’t simply make everything quicker!

Sure, businesses, especially those directly involved in tech, may be able to keep up with the pace of change… for a time. Because, at some point (even if it is indirect) all businesses rely on education, social, and governmental institutions to function. And, let’s not forget, if people lose jobs and earning power, there will be no one to buy from businesses.

But, even beyond that, there’s a bigger, truly fundament point that is as profound as it is simple. It was, perhaps, best put by the great HG Wells, when he said “If anything is possible, then nothing is interesting.” This, I would argue is one of the most serious problems of AI. We are, in our pursuit of technological process, giving little thought to its effect on the very base motivations for life. If we can all create everything, and we create more than anyone can possibly take the time to appreciate, then why bother creating anything?

And, before someone pipes up with “You sound like those people who thought riding on a steam train at 30 mph would make you suffocate due to lack of oxygen”, context is everything. There’s nothing wrong with AI itself, in isolation. It is simply the rate of change it manifests that is outstripping our natural capacity to adapt. Maybe this rate of change wouldn’t be a problem if we were some pan-galactic star-faring civilsation, or if there were only a few million people on Earth, and we had to be vastly more productive each. But, even in those scenarios, I suspect it wouldn’t be long before AI simply rendered humans obsolete and we just faded into the background.

So, in conclusion, we must stop comparing the AI revolution to past technological changes, and discuss it in the context of reality, and our human biological and sociological constraints, which are very real. Only then can we truly see the dangers of AI.