Will ChatGPT herald the next development epoch?

Firstly, we must understand how we got here, before asking what’s next:

There were, broadly, roughly eight major turning points in that time. Most of them didn’t involve any fundamentally new idea – they just used the advancements going on in the background to make a new thing work. And since those advancements became more and more enabled by each progressive step, so too were we able to make more and more things work.

The first was language, around 60,000 to 100,000 years ago. Language allowed humans to begin preserving knowledge from generation to generation, meaning that each generation didn’t have to relearn everything. Some other animals can pass on a few things – other primates teach their children how to use tools, for example – but not with the speed or detail allowed by human language. It’s very possible that many people independently invented the idea of language, but – well, if you have an idea of language and other people don’t, how do you tell them about it?

The second was agriculture, around 10,000 years ago. Before the rise of agriculture, humans spent most of their activity just getting enough food to live. After it, you could spare enough resources to have people start to become experts in things. One person could become an expert at computing things, another could become an expert tailor, and so on. And those experts pushed the boundaries of their fields forward (and shared what they learned through language). Agriculture was developed simultaneously around the world at about the same time, which suggests there was some underlying reason it happened then – the most obvious explanation is that this coincides with Earth’s climate warming up after the last glacial period. The idea of planting things probably wasn’t new, but this is the point where you could actually settle down with a field of crops.

The third was writing, around 5,500 years ago (possibly earlier, but cuneiform is the oldest we know of). Writing was a huge step forward. Now you could pass information on without having to have an expert sit down and tell you. You could keep track of things for generations, and find patterns in what you kept track of. (For example, a lot of early mathematics was worked out to predict celestial events like eclipses, which you could only do if you had reliable records of when eclipses had occurred in the past.) Writing also arose independently in many different places, but only in places that had settled civilizations, suggesting that settling down was probably a necessary prerequisite to figure it out (or at least to keep documents that survived long enough to be discovered). Marking things wasn’t new, but someone had to figure out a system that could be learned and shared.

The fourth was the printing press, around 600 years ago. This kicked writing into high gear and made information orders of magnitude more available (the printing press could produce copies of text around 100x faster than previous printing methods). While I didn’t include metallurgy in my list here, it was a very important field throughout ancient and pre-modern history, and between the previous bullet and this one, you went through the Bronze and Iron ages. And it was knowledge of metallurgy that made the printing press possible. Gutenberg was a metal-smith and one of his big innovations was using a new alloy of metals that made the physical mechanics of the printing press work. The idea wasn’t new, but Gutenberg built one that actually worked.

The fifth was the Scientific Method, around 400 years ago. While earlier scientists (as we’d call them today) had existed, the scientific method turns out to be a much more effective way to test ideas than most previous frameworks, which tended to start with logic and try to explain observations, rather than observing observations and trying to design theories that fit them. Again, the idea of studying the world wasn’t new, but a particular approach that happened to work better than previous ones allowed progress.

The sixth was the factory and mass production, around 150 years ago, which was made possible by massive advances in chemistry, engineering, and materials science enabled by the scientific method. This made the tools and apparatus for experiments, data collection, and observation far more available, not just things that could be afforded by a few very wealthy researchers (or those funded by wealthy benefactors). Automation wasn’t new, but advancements in things like metallurgy, steam power, and later electricity made automation work in a way the machines of the ancient world didn’t.

The seventh was electronics, which was spread throughout the 20th century. The rise of electronic machinery allowed a whole new range of observations and a new level of precision. Again, this was only possible because factories permitted the mass-production of electronic components and because materials-science had advanced to the point that things like the transistor – a key component of all modern electronics that makes logical circuits possible – could be produced.

And the eighth was networking, currently in the form of the Internet on which we are having this conversation. The idea of a network, of course, was not new. Networks of information date back to the ancient world. But advancements in speed and precision via electronics, and the creation of vast networks of infrastructure through mass production, made it possible to automatically shoot vast amounts of information around the world on demand.

In general, it’s best not to think of things in terms of “one big idea”. Big ideas are enabled by a million small advancements, and many big ideas have been had a million times before someone turns the big idea into a working thing. And the pace gets set by those small advancements, not by how many big ideas get had.

Is ChatGPT the next major advancement? Probably not. Remember GPT is not AI – in fact, it is non-intelligent — it is merely a method of organising words in a contextually correct order. But, it may well allow intelligent humans the time and freedom to start conceptualising the next major advancement.