Civilization from scratch

This post is not about AI and not about winter. I have a few of those coming, but this one is about something different. I hope you don't mind.

A friend of mine recently gave a lot to think about by stating the following thought experiment:

Imagine you are taken back in time. To what extent would you be able to advance the civilization of the given era with all the knowledge in your head (no notebooks).

Initially the reaction is obviously that since we all live and breathe the current technical civilization, one should be able to recover almost everything right? There are some many uncertainties to which we already know the answers, so this should be much easier than to get there without such insight?

When you actually give some thought to it, you will realize that things may not be so easy. First of all, in most cases if somebody was taken back in time but left in the same place, they would end up in a middle of nowhere and would have to first survive to even get into contact with any contemporary humans. Say San Diego 300 years ago was an empty costal desert, and I would probably die within several days of landing here back then. Most of the modern day people in such situation would have starved to death or got eaten by predators.

For those who would survive this initial threat, the next major challenge would be to contact the humans of the era. Putting aside the obvious communication problems, many people in ancient times would not be very welcoming to strangers, particularly weird strangers. Chances are good fraction of such encounters would result in a homicide.

Those lucky enough to pass this first encounter would have to merge into the social structure, and even with all the knowledge of the wonders of the future, the initial social allocation would be at the very low end. With scarce resources, societies of the past were rather cruel and there was no social safety net. So chances are, for the first few months of your existence in the past you'd have to work hard to support yourself and earn any social privileges. In the meanwhile you'd not be able to introduce any brilliant ideas about how to improve the existence because frankly nobody would trust you and you'd be just some weirdo a the very low end of the social order.

If all went well, after some time you'd finally climb the social ladder high enough to enjoy some privileges. If things were really lucky you could get a position of an alchemist at some castle or an independent philosopher if you ended up in the ancient Greece. Either way, let's assume you were actually socially stabilized and wealthy enough to be able to afford most of what was available in a given epoch and people actually would listen to you and trust you. What would be the lowest hanging fruits? The bits of information that could really change the way people lived in the past?

My friend is an immunologist, and hence his first choice was the introduction of hygiene - the idea that disease is spread by invisible germs and that washing hands and food with fresh water and using boiled water for drinking eliminates most of the exposure.

This sounds really easy and indeed life changing, however there would be little proof available that this really works and in many epochs such as middle age, this would be a clear blasphemy and could land you in jail (or more likely executed). So as absurd as this sounds, even such an obvious and easy improvement would have to be sold really carefully.

I've been thinking about other things constantly over the past few weeks and noticed how surprisingly difficult would it be to reconstruct any meaningful part of our modern-day civilization. One would think - a steam engine!

Yes the steam engine is a great idea, but note that without any machinery for shaping metal and without modern metallurgy even that would be quite difficult to pull off. Obtaining the proper material and shaping it into cylinder would be very expensive and error prone. However at least with the steam machine it is reasonably simple that most people could recover the basics of the construction from memory (remember, the essence of this experiment is that you are left there only with what you have in memory, no books, no manuals).

Anyway, before steam engine would become reality you'd have to at least partially recover some of the industrial processes for metalworking, such as casting, hardening, shaping. You'd need to build a primitive lathe, you'd have to be able to come up with a cutting tool of higher hardness than the metal you'd try to cut. Do you know how to harden steel? How to make steel in the first place?

Building reliable steam engines is a possibly lost craft.

Even if you could indeed build a simple steam engine, it would take a lot more time to employ it to things such as trains or production lines. When you look closely at a 100 year old steam engine (see pic above), you will notice lot's of details, many of which were likely important to make this machine reliable. The way things are mounted, valves controlled, lubrication provided. All of these details are a (potentially lost) craft which took many decades to figure out.

Next thing that comes to mind is electricity. But this is a can of worms too. First of all, maybe you could build a battery, using some copper and zinc and some acidic solution. But next you'd need to get some wire, and wire was not readily available in pre-industrial times. You'd have to forge some wire yourself, and although certainly possible, would be a time consuming process.

Once you had a wire, you could build a telegraph. That would be a big deal, all you need is a long wire and some coils and the battery (and reinvent the Morse code). I think this would be quite doable, though depending on the epoch you were stuck in, would be very expensive. But if you could demonstrate long range communication abilities, that would give a huge strategic advances and could win you a lot of kudos.

Batteries would be maybe enough for a simple telegraph, but certainly not for any more demanding uses of electricity. If you had a lot of wire you could try building an electric motor/generator. It is relatively easy to build a motor if you had a permanent magnet available, but magnets were not generally available in pre industrial times, though if you were lucky and rich you could get hold of a lodestone. Even if you had a piece of magnetic iron and some copper wire, you'd still need to create a rather precise shaft and some form of a bearing, do the proper wiring and some form of brushes to connect everything up. All of the above seem simple in principle, but if you are not actively machining metal, you may find that making any of these parts could be very time consuming and frustrating. It is hard to appreciate how much craftsmanship is necessary to obtain what we take for granted today. Clickspring youtube channel is one of several in which those matters are being considered in the context of reconstructing the Antikythera meachnism. Long story short, to get any precise tooling, and get these things right is a hell lot of work and hands on knowledge.

Having electricity generated in more appreciable quantities is the basic requirement, since good 95% of our modern industrial civilization runs on electricity. Having this, one could start thinking about creating a lightbulb, though this would be a challenge on its own - getting the filament, getting the glass vessel, shaping everything together and above everything else - evacuating the air out. Each one of these steps would be a multi-month long project and that is assuming you know what you are doing. And frankly? I don't think average person taken off the street today would be able to get this far anyway.

Which brings me to the crux of this post: how much do we actually know? I have a friend, an electrical engineer. I once asked him some question about the electrical wiring of a typical home, something about the difference between the non-hot wire and the ground or something. And he laughed at me and said he is not an electrician but an electric engineer so he does not know. I'm sure he could figure it out, but this illustrates the central point - we only know enough, to move the civilization from where it is today a tiny step further, but we generally assume everything else is given. We don't generally have the complete knowledge to recover everything from scratch.

We are standing on the shoulders of giants.

The reader obviously may have different expertise and choose different things. E.g. somebody who knows how to synthesize penicillin could certainly leverage that skill to everybody's benefit. I, out of the top of my head would not be able to figure that out probably. Along with a million other things that just seem so obvious we don't even care. Basically most of my technical expertise in computer science/computer vision/machine learning would have been completely useless in the past. Particularly anytime before XX century.

We rely on books and expertise of others, and a lot of that expertise is not even expressed in books, but rather carried from generation to generation through apprenticeship as hands on experience (actually I suspect way more than we would imagine is transferred this way). We rely that somebody out there will know how to dig the ore, somebody else will know how to purify it, someone will melt it into metal, someone will machine it into shape, someone will put the parts together and so on and so on.

The technical knowledge is part of our culture. It is carried by books, by hands on skills, by what we could call a "knowhow". It is more fragile than one might think - missing either of the carriers - verbal knowledge expressed in books and the non verbal knowledge carried in the skills of individuals would cripple our technical status-quo, much like the legendary fire of the library of Alexandria is believed to have crippled ancient scientific achievements. Perhaps this helps explain why 50 years after sending man to the Moon, we appear not be technically capable to repeat that achievement, even though it was very well documented.

If you found an error, highlight it and press Shift + Enter or click here to inform us.