Obviously something of this magnitude will have blindspots. This tech tree seems to be vastly underselling the impact of advances in metallurgy and precision machining. As well as most of what you might call "basic science".
This leads to e.g. the Gas Turbine just appearing out of nowhere, not depending on any previous technology
They tried to define what they mean by technology [1], but they seemingly gave up on it partway through. Had they followed it consistently, they would have excluded certain cultural-practice-based technologies like nixtamalization that made the list.
The inconsistent definition and the pretty large gaps leads to a lot of oddness. Just look at how sparse anything related to textiles is. "Clothing" just gets one "invention" in 168k B.P., even though a t-shirt and an arctic jacket are obviously very different technologies. New world agriculture is similarly strange. Nodes appear from nowhere and lead nowhere, presumably because there are implicit "nature" edges they didn't want to represent as technology.
Feel like if you're doing something like this you should just basically maximalize your definition. The fun here is seeing all the nodes, obviously!
Maybe then you get into arguments about whether the dependencies were "required", but there it's more or less resolvable by relying on what "actually" happened rather than the minimal tree (which is its own exercise)
You will end up writing a technology focused complete history of the world. That tapestry would be absolutely fascinating to be able to navigate and it would also be always out of date.
Yes it feels very 'male' centric in a way. Like with clothing, the various stitches that you need to make textiles, each individual stitch method is it's own technology. Like, every knot that gets invented, that's a technology. Knots and stitches and warp and weft, they can only come out of a human mind, they are inventions. But trying to find out the dates there, that's nearly impossible if the data were recorded even. And then most of those textile inventions were done by women who are then again historical dark matter even in the best circumstances.
Still, great project and I'd have loved to see it crowdsourced like wikipedia.
> Had they followed it consistently, they would have excluded certain cultural-practice-based technologies like nixtamalization that made the list.
This is an interesting example. It's a technology that's very important for staying alive, but not one that you'd expect to contribute to any kind of progress. It's just something you have to do to corn before eating it.
I'm a former archaeologist, so my personal definition of technology is extremely expansive.
You don't actually need to nixtamalize maize. It's totally edible without and most americans today don't eat nixtamalized corn outside masa. It's just a process to make it more nutritious and importantly, nearly nutritionally complete. For ancient societies, nixtamalizing had a role similar to things like vaccination do for us today. It reduced malnutrition and the economic/social/political effects of disease. The difference I'm trying to highlight is that it probably wasn't understood as such and intentionally done for that purpose. Nixtamalization was culturally encoded as just what you did. Had they had a better understanding of nutrition, they probably would have made more intentional efforts to include the missing vitamins nixtamalization doesn't provide. We often see signs of those missing nutrients in precolumbian skeletons.
This extends to a surprisingly wide variety of ancient technology. Most metallurgy probably wasn't understood in the technical sense we think of it today until quite late. We see that with early glass, where people simply didn't understand what they were doing. Ingredients from specific areas would have specific effects, but sometimes didn't for reasons no one at the time understood. Craft communities would standardize on very specific, ritualized processes that simply couldn't be changed because they didn't have a good mechanistic understanding of the variables involved. One of the downstream effects of this is that poaching craftspeople is a viable strategy (they had the specific "recipe") and also that resources like sand from specific areas in syria and egypt were effectively non-fungible for centuries. You had to trade with whoever controlled that area even if you had the craftspeople.
Craftsmen having secret recipes isn’t quite restricted to the past. According to my drummer friends, nobody knew how to make cymbals as good as Zildjian’s until very, very recently.
A lot of tradecraft is in the most literal sense tradition, passed from one person to another with no other context than 'it works' and 'this is the way it is done'. The analogies in the software world are probably libraries, design patterns and languages. Each of these embody a lot of knowledge without being - usually - very explicit about the reasons why they are the way they are.
Researching those reasons requires a lot of work and most people just want to get to the result step without necessarily understanding why you put the oil in the pan before the egg. They're hungry.
Andrew Carnegie wrote that one of the things that gave him an advantage over other steel manufacturers was that he hired a chemist to test ore for iron content.
By implication, this was something that had never been done before.
I don't think you're right about that. They judged ore, as the comment above mentions, by the location it came from. They tried standard techniques to process whatever ore they had, and if those techniques led to a bad result, they considered the ore bad.
Andrew Carnegie discovered that certain "bad" ore was that way because its iron content was much higher than usual. Secure in the knowledge that this ore was actually better than "good" ore, he developed the techniques to use it.
On the other hand, it’s basically impossible to have large organized civilization in Mesoamerica without hominy[1]. So should hominy be upstream of anything Mesoamericans invented?
1: For anyone whose Nahuatl is a bit rusty: the English word for nixtmalized corn is “hominy”.
Almost all science is tightly bound with advances in material science, often the driving each other in alternating steps like interlocking gears. One engine driving those gears is war, another is population growth and education.
There are obvious exceptions, such as Math, Philosophy (insert all links lead to philosophy here). But even Math is seeing progress in materials science as a component now (computer derived proofs for instance).
Making a really good tech tree is a stupendously hard problem. I once started working on one for a game but gave up once I realized that doing this properly is probably going to take a lifetime or two and there are other things I can do that are more immediately useful.
Even aside from the big advances, look at things like clothing in the past few decades. Outdoor clothing has changed enormously but so have so many other day-to-day things--a lot of which are about electronics but that's materials-related as well.
It's an interesting question. Why couldn't the Romans have invented $X? And the answer is mostly the tech tree. There are probably exceptions around things like germ theory of disease and so forth but it's mostly true.
There is no 'most important' human technology. All of it interlocks, and usually the prerequisite steps all need to be followed before you can progress to the next level. I wonder how long it would take given a paper copy of wikipedia (hopefully printed on acid free paper) to get back to a functional technology society. I'm sure it would go faster than the first time around, but I'm not so sure it would be less than a few hundred years.
A lot of those things are incremental improvements that build onto each other, like refining an alloy by a few % many times over to end up with something entirely different.
How would one determine what is sufficiently different to deserve a node?
But 100% agree, incremental improvements are the vast majority of advances.
Unfortunately, a lot of the reasoning behind omissions is that we have lost the majority of information; what we learned was from luck, inference, and unfortunately (because it produces misleading results), selective durability of some items over others.
My particular interest is in screw cutting lathes, and it appears that the Wikipedia entry[1] (on which this seems to be based) was off by about 25 years (1775 instead of 1800), and thus copied to this work. I've let the folks at Wikipedia know.
Interesting. On that note, Da Vinci's design (which I was fortunate enough to see a replica of at a local museum) was also very clever, being suited not only for screw cutting but also screw origination, as it could make new screws more accurately than the two leadscrews in the machine itself, and swap them out to improve its own accuracy. But I suppose it doesn't extend that date even further back because it wasn't a general purpose lathe, it could only cut screws.
The ultimate reference for this is the Foundations of Mechanical Accuracy, which is available as a PDF[1], the print versions aren't cheap.
One example of precision is the Whitworth 3 plate method (invented in the 1830s), with this method, you can make surfaces that approach precision in the micron range, if necessary.
Leonardo Da Vinici's screw cutting lathe sketches show a machine that uses 2 lead screws to generate another, which averages the values of the other two screws, using careful rotations of threads, etc, it should be possible to use this method to work your way up to a uniform precise screw where each new generation is better than the ones before it.
The main way is by averaging error terms toward zero.
For example, Da Vinci's machine cuts the new screw blank in the centre of the carriage, which is driven by leadscrews on either side. The nuts would have likely been something like a wide strip of leather clamped over the screw thread, so there's enough compliance to average over a few pitches of thread, and the position of the cutter would be close to the average of the position set by each leadscrew thread.
Imagine the rough-cut screw threads have a pitch vs rotation angle described by p1(θ) and p2(θ). Running the machine then creates a new screw which is nearly a duplicate of the drive-screws in the machine but with pitch p3(θ) = (p1(θ) + p2(θ))/2. You can make two of these screws and swap them for the two leadscrews in the machine (it's built to be easy to do this). The random errors from a rough-cut screw gradually average out. But the cleverness doesn't end there. You then flip one of the screws backward end-to-end, so now you're averaging p3(θ) with p4(L-θ). You can also offset θ by any amount for either screw by offsetting the change-gear and re-clamping the carriage nut. Repeating these actions, you gradually can eliminate all systematic thread errors from the initial rough-cut screws and converge towards cutting a screw with nearly-constant pitch.
(It doesn't end there either; there's really a lot of flexibility with Da Vinci's design. Changing the gear ratio lets you create a fine-pitch thread from a coarse-pitch thread, or vise-versa, or cut a multi-start screw by rotating the blank 180 degrees or end-over-end).
The traditional way of making a multi-start screw is to use a dividing head and rotate the part an appropriate angle. If you're working with change gears, choosing a gear that's a multiple of the number of starts could provide alternate way to do this.
Oh yes, flipping the screw blank end-over-end would be an awkward way to do it unless you wanted to make a screw with two starts each having opposite errors. Once the screws have gotten good enough you wouldn't need to bother with things like that.
Just some examples: take a string, don't bother to measure it, just any length between 1 and two meters or so would do. Take a pencil (or a piece of charcoal if you really want to go native) and a smooth branch. Stick the branch in the ground, tie the string around it so that it can slide with little friction and put the pencil in a loop of string on the other side. Now use this to create a circle. You started off with very rough elements not specifically sized for any purpose and ended up with a high precision representation of a mathematical concept.
Another: take a bunch of roughly cast metal balls. Put them on a sieve and let it vibrate until the balls have all passed through the holes in the sieve. Behold: metal spheres, so precise that you probably can't really measure the degree to which they are not spherical without resorting to instruments that you're not supposed to have in this scenario. Then sort by weight (which is a proxy for size). Now you can make ball bearings.
I find this video of theirs the most relevant [0] where they go through how to start on a desert island and build a flat reference plate using the three plate method and the build up from there
Have you ever thought about how alien lifeforms would probably invent screw cutting lathes too? The screw feels like such a "human thing", but what else would serve the rotational wedging purpose in this universe's elements and physics?
It would be funny to think of what might actually be a human thing. Like, our arms are quite weird, right? So potentially another intelligent species even on a rocky planet might not invent throwing spears, right? Even our close relatives, the chimpanzees, can’t use them well at all. Even fairly flighty animals seem to barely have the idea that a creature standing, like, tens of feet away from them might be “in range.”
I wonder what the trajectory (no pun intended) of the development of melee spears would be, without throwing spears.
The spear being a wildly popular a successful weapon for almost all of history, any changes to spears would, I guess, make a big difference.
This 2D map is hard to explore since it's so sparse. I have to follow lines to find each thing, since it's 99% empty void. Is there a snap to next item hotkey? Am I just doing it wrong?
There is a "jump to nearest" button that appears if nothing is visible, but maybe it's too subtle? A hotkey would be a good idea too. I find that the best way to navigate is the minimap
Additionally I've always wanted institutions to be part of the timeline of technology. Corporations, Nation-states, Universities, Guilds, International Organizations - the ways people innovatively organize make things possible that otherwise wouldn't be.
The higgs boson experiments, for example wouldn't have been possible without the complex international institutions that orchestrated it. Manhattan project, Moon landing, the internet ... the iphone ...
Found a minor nit — the entry for “shoe” is at 3500 BCE and links to the Wikipedia article for “sandal”. But the Wikipedia article for “shoe” [1] indicates that the earliest shoes (apparently not sandals) were worn around 7000-8000 BCE.
Whether a node is a terminus (or root without predecessors) is basically ~never meaningful in the tree, it's almost always just missing data. Here it seems pretty clear I omitted a link from Chinese writing to woodblock printing. Fixed!
(Not an expert but) Chinese writing is hieroglyphic, wheres 'western writing' is phonetic. Western writing has a very small character set and is thus well-suited to a printing press, whereas hieroglyphics have thousands of characters (for thousands of concepts) but aren't fundamentally linked to the language like western characters are.
> Western writing has a very small character set and is thus well-suited to a printing press
This is vastly overstated. This was a widely popularized idea in the west but has largely been debunked by more recent scholarship that is less interested in demonstrating the superiority of the west.
Tom Mullaney at Stanford has a good book on chinese typewriters (this narrative is usually presented against both printing presses and typewriters) and is a good entrance into the intersection of chinese script and technology.
History research is typically published in books rather than papers, so it isn't content I can link to directly.
upstream of what? Did you not look for mesopotamian or egyptian writing, which predate chinese writing by thousands of years. This sounds more like a chinese bias lol.
But they won't do it if they don't like this sort of approach! Which I think is wrong of them btw. But I don't think I've heard from historians much since this project went public.
Surely you'd understand how this sort of approach eliminates all contingency from the past. Especially given that tech trees are usually present in video games where they look forward rather than backward.
If you are doing a large project on the history of science (you have clearly put a ton of work into this) but you aren't communicating with historians of science that's concerning to me.
If you like this then you will probably enjoy the book How to Invent Everything: A Survival Guide for the Stranded Time Traveler by cartoonist and computer-scientist Ryan North.
It's interesting that prior to the industrial revolution there are still some periods where it seems like innovations arrived relatively fast, and others where it was comparatively slow. E.g. a lot more entries are in the 500 BCE - 200 BCE period than the 200 - 500 range.
Although the idea of a "Dark Age" is mostly debunked these days, the slow unraveling of the Western Roman Empire led to a real and sustained change in material conditions. Notably, population density and urbanization both decreased, along with the labor specialization that accompanies them. I'd expect most 'inventions' to happen when and where people have the most hands on time to make them! (I can't really speak to Indian and Chinese civilizations, but they have also had integration and disintegration periods)
I wonder if something similar could be added here where I say something like "what's the most important descendant of x" and it would bring me to that tech and give me a little explanation of why
Somewhat tangential, but can anyone explain or give any insight into how candles were invented? This has always puzzled me. The idea that if you put a piece of string in some wax, and light it on fire, the string will get hot enough to melt wax, and the molten wax will be sucked into it by capillary action, and burn, before the string itself burns, seems very counterintuitive and hard to come up with before a modern understanding of chemistry and physics. And yet, candles have apparently existed since 700 BC. How?
Pure speculation, but I would guess people first mixed wax and lamp oil in different ways to still get the burning effect of oil, with less of the cost of the oil, then added a wick to help light the oil/wax.
Then eventually that product morphed over time to the point where they realized the oil wasn't actually a necessary component
Good idea, it sounds plausible! But it still leaves open the question of how oil lamps were invented. How did someone figure out that a wick would be helpful?
Rope was around long before the wick. It seems conceivable that rope shavings or pieces or old rope were an easy way to start a fire.
This was then used with oil to make an even better fire starter or means of transferring fire. Eventually someone realises that a rope soaking in oil is easily lit and sustains a flame.
Before wicks, how do you burn oil? It's not easy to just ignite a bucket of lamp oil (putting aside what you might make the bucket out of). Probably you soak other fuel like wood or rags in the oil and burn the result. It's not a huge step from there to accidentally find out that you can make do with one piece of wood or cloth or string for a lot of oil, assuming you have something to put the oil in.
1. Someone dips a rope in flammable oil before lighting it, and sees that it's quite flammable.
2. Some other time, someone tries to use a rope dipped in flammable oil as a fuse to trigger the lighting of the oil once it burns back to the oil
3. They notice that the fuse keeps burning but doesn't burn back - in other words, the wicking effect
4. They shorten the rope and reshape the pot, and that's an oil lamp.
The first oil-lamps were basically a bowl with animal-fat and plant parts as wick. It's not hard to imagine how ten thousands years ago a hunting tribe could discover such a device by accident or on purpose.
Because the people from 700 BC were just as smart as we are. Observing the world around you is one of the best sources for ideas and there are predecessors to 'real' candles that must have fertilized the ground for the discovery. For instance, a stick dipped in rendered fat could serve as a torch. Not quite a candle but a significant step on the way there.
It's funny that there are so many innovations right now the recent part of the chart just has to arbitrarily exclude an insane amount of stuff innovation that's happening.
No HIV vaccine. mRNA vaccine get's a single entry instead of vaccine per disease like prior vaccines. No battery stuff since 1985. Just amazing, fractal improvement is everywhere.
Great phrase - fractal improvement. It's kind of the idea of this book [0]
Even more cool: commercial progress trails tech. It takes a long time for companies to figure out how to turn a new idea or a cheaper input into a new product/industry, and then for related companies to grow into an economic ecosystem.
So one would expect to see some spectacular economics over the next couple of centuries.
Beautiful! I wonder if Jimmy Maher's heard about this; he wanted something like it for The Analog Antiquarian back ages ago before he kicked that off, as a way of reflecting the span of history in the structure of the index/TOC, but we never could figure out really how to get it to go anywhere we liked. It's a surprisingly tricky problem, and this is an impressive realization!
This is cool, but I think the execution is off because there's so much empty space. I think it would work better if the nodes were much smaller and closer together so you can see more of the graph in one screen.
Genuine question: how does it make sense that both bread and alcohol fermentation were invented before the domestication of any grain? Were people making bread (and beer!) with gathered seeds?
Yes, there are archeological remains of bread made from wild wheat and barley by hunter-gatherers in Jordan. It makes sense when you think about it, domestication would have happened only if the wild plant was useful, and if the plant was useful then it's likely people experimented with various ways of using it (like baking)
Wine is much easier than beer, because ripe fruit will ferment on it's own. Even wild animals get drunk sometimes; it's sometimes said that wine was discovered but beer was invented.
So alcohol without any grain is easy, but I don't know the answer for beer. On the other hand, why would you domesticate grain unless you already knew you could turn it into beer or bread?
Grain can be turned into porridge (so basically just soaked) and eaten, bread and beer is optional. In fact, IIRC classical-era armies would eat their grain as porridge instead of spending hours in camp baking it into bread. For obvious reasons.
Tobacco and alcohol are included, but there is no mention of other botanical or mycological developments outside food production. Those would be an interesting addition.
Highly recommend the Dr. Stone anime if you're interested in a story with the premise of starting civilization from scratch but armed with the sum total of modern human knowledge about science and engineering.
Also, if you want even further back precedent for this kind of plot device, I highly recommend reading a Connecticut Yankee in King Arthur's court by Mark Twain.
I second this. It's the only show I've seen making a semi-realistic attempt at this (ignoring the absurdity of the initial petrification in the first place and Dr. Stone having superhuman knowledge of all human inventions)
I'd also recommend the Destiny's Crucible series - the basic premise is that a chemist from our world is transported to another planet of humans at a much lower technological level, and some moderately standard isekai hijinks ensue.
I read five of the books, and really enjoyed them; if you like the "competence porn" genre of novels, this is a pretty good one.
See... now, I love that type of show/comic/book/etc. And now that I have a name for it, I want to search for more. But I very much do _not_ want to search for that term. Lol
I watch this with my daughter and we love it. I love shows with "narration", talking about the context/details of things, and Dr Stone really nails that (I know the main character isn't really a narrator.. but it accomplishes the same thing).
Its a great start! Bound to have bias and blindspots. It would be cool to run an agent that could incrementally enrich this knowledge graph. Take some modern day technologies and backtrace the components and their development.
Obviously something of this magnitude will have blindspots. This tech tree seems to be vastly underselling the impact of advances in metallurgy and precision machining. As well as most of what you might call "basic science".
This leads to e.g. the Gas Turbine just appearing out of nowhere, not depending on any previous technology
They tried to define what they mean by technology [1], but they seemingly gave up on it partway through. Had they followed it consistently, they would have excluded certain cultural-practice-based technologies like nixtamalization that made the list.
The inconsistent definition and the pretty large gaps leads to a lot of oddness. Just look at how sparse anything related to textiles is. "Clothing" just gets one "invention" in 168k B.P., even though a t-shirt and an arctic jacket are obviously very different technologies. New world agriculture is similarly strange. Nodes appear from nowhere and lead nowhere, presumably because there are implicit "nature" edges they didn't want to represent as technology.
[1] https://www.hopefulmons.com/p/what-counts-as-a-technology
Feel like if you're doing something like this you should just basically maximalize your definition. The fun here is seeing all the nodes, obviously!
Maybe then you get into arguments about whether the dependencies were "required", but there it's more or less resolvable by relying on what "actually" happened rather than the minimal tree (which is its own exercise)
You will end up writing a technology focused complete history of the world. That tapestry would be absolutely fascinating to be able to navigate and it would also be always out of date.
Yes it feels very 'male' centric in a way. Like with clothing, the various stitches that you need to make textiles, each individual stitch method is it's own technology. Like, every knot that gets invented, that's a technology. Knots and stitches and warp and weft, they can only come out of a human mind, they are inventions. But trying to find out the dates there, that's nearly impossible if the data were recorded even. And then most of those textile inventions were done by women who are then again historical dark matter even in the best circumstances.
Still, great project and I'd have loved to see it crowdsourced like wikipedia.
> Had they followed it consistently, they would have excluded certain cultural-practice-based technologies like nixtamalization that made the list.
This is an interesting example. It's a technology that's very important for staying alive, but not one that you'd expect to contribute to any kind of progress. It's just something you have to do to corn before eating it.
I'm a former archaeologist, so my personal definition of technology is extremely expansive.
You don't actually need to nixtamalize maize. It's totally edible without and most americans today don't eat nixtamalized corn outside masa. It's just a process to make it more nutritious and importantly, nearly nutritionally complete. For ancient societies, nixtamalizing had a role similar to things like vaccination do for us today. It reduced malnutrition and the economic/social/political effects of disease. The difference I'm trying to highlight is that it probably wasn't understood as such and intentionally done for that purpose. Nixtamalization was culturally encoded as just what you did. Had they had a better understanding of nutrition, they probably would have made more intentional efforts to include the missing vitamins nixtamalization doesn't provide. We often see signs of those missing nutrients in precolumbian skeletons.
This extends to a surprisingly wide variety of ancient technology. Most metallurgy probably wasn't understood in the technical sense we think of it today until quite late. We see that with early glass, where people simply didn't understand what they were doing. Ingredients from specific areas would have specific effects, but sometimes didn't for reasons no one at the time understood. Craft communities would standardize on very specific, ritualized processes that simply couldn't be changed because they didn't have a good mechanistic understanding of the variables involved. One of the downstream effects of this is that poaching craftspeople is a viable strategy (they had the specific "recipe") and also that resources like sand from specific areas in syria and egypt were effectively non-fungible for centuries. You had to trade with whoever controlled that area even if you had the craftspeople.
Craftsmen having secret recipes isn’t quite restricted to the past. According to my drummer friends, nobody knew how to make cymbals as good as Zildjian’s until very, very recently.
A lot of tradecraft is in the most literal sense tradition, passed from one person to another with no other context than 'it works' and 'this is the way it is done'. The analogies in the software world are probably libraries, design patterns and languages. Each of these embody a lot of knowledge without being - usually - very explicit about the reasons why they are the way they are.
Researching those reasons requires a lot of work and most people just want to get to the result step without necessarily understanding why you put the oil in the pan before the egg. They're hungry.
Andrew Carnegie wrote that one of the things that gave him an advantage over other steel manufacturers was that he hired a chemist to test ore for iron content.
By implication, this was something that had never been done before.
I just looked it up, and he was born the year after Mendeleev, who'd go on to design/arrange the periodic table of elements.
I'm guessing they had traditional assaying techniques, just with less accuracy than a contemporary chemist.
I don't think you're right about that. They judged ore, as the comment above mentions, by the location it came from. They tried standard techniques to process whatever ore they had, and if those techniques led to a bad result, they considered the ore bad.
Andrew Carnegie discovered that certain "bad" ore was that way because its iron content was much higher than usual. Secure in the knowledge that this ore was actually better than "good" ore, he developed the techniques to use it.
On the other hand, it’s basically impossible to have large organized civilization in Mesoamerica without hominy[1]. So should hominy be upstream of anything Mesoamericans invented?
1: For anyone whose Nahuatl is a bit rusty: the English word for nixtmalized corn is “hominy”.
Almost all science is tightly bound with advances in material science, often the driving each other in alternating steps like interlocking gears. One engine driving those gears is war, another is population growth and education.
There are obvious exceptions, such as Math, Philosophy (insert all links lead to philosophy here). But even Math is seeing progress in materials science as a component now (computer derived proofs for instance).
Making a really good tech tree is a stupendously hard problem. I once started working on one for a game but gave up once I realized that doing this properly is probably going to take a lifetime or two and there are other things I can do that are more immediately useful.
Even aside from the big advances, look at things like clothing in the past few decades. Outdoor clothing has changed enormously but so have so many other day-to-day things--a lot of which are about electronics but that's materials-related as well.
It's an interesting question. Why couldn't the Romans have invented $X? And the answer is mostly the tech tree. There are probably exceptions around things like germ theory of disease and so forth but it's mostly true.
Any post about tech trees sooner or later mentions 'I, pencil', so let's get it over with:
https://fee.org/ebooks/i-pencil/
They are expecting suggestions for this work in progress.
https://www.historicaltechtree.com/about#contributing
Metallurgy is by far the most important human technology. That's why "copper", "bronze" and "iron" are used to classify ages.
There is no 'most important' human technology. All of it interlocks, and usually the prerequisite steps all need to be followed before you can progress to the next level. I wonder how long it would take given a paper copy of wikipedia (hopefully printed on acid free paper) to get back to a functional technology society. I'm sure it would go faster than the first time around, but I'm not so sure it would be less than a few hundred years.
Wikipedia is useless for rebooting technological society, it doesn't show where or how to mine the necessary minerals.
Even more than agriculture?
Metal has the fortunate process of durability over millenia... we are biased by what we find.
A lot of those things are incremental improvements that build onto each other, like refining an alloy by a few % many times over to end up with something entirely different.
How would one determine what is sufficiently different to deserve a node?
But 100% agree, incremental improvements are the vast majority of advances.
Unfortunately, a lot of the reasoning behind omissions is that we have lost the majority of information; what we learned was from luck, inference, and unfortunately (because it produces misleading results), selective durability of some items over others.
[dead]
Discussed once (and I do mean once):
Historical Tech Tree - https://news.ycombinator.com/item?id=44104243 - May 2025 (1 comment)
Edit: oops!
The Universal Tech Tree - https://news.ycombinator.com/item?id=44161607 - June 2025 (65 comments)
Someone else found https://news.ycombinator.com/item?id=44161607
(A link to an article about how it's made, with 65 comments)
Oh wow, thanks for letting me know! Added above.
My particular interest is in screw cutting lathes, and it appears that the Wikipedia entry[1] (on which this seems to be based) was off by about 25 years (1775 instead of 1800), and thus copied to this work. I've let the folks at Wikipedia know.
[1] https://en.wikipedia.org/wiki/Screw-cutting_lathe
Interesting. On that note, Da Vinci's design (which I was fortunate enough to see a replica of at a local museum) was also very clever, being suited not only for screw cutting but also screw origination, as it could make new screws more accurately than the two leadscrews in the machine itself, and swap them out to improve its own accuracy. But I suppose it doesn't extend that date even further back because it wasn't a general purpose lathe, it could only cut screws.
having no experience in precision manufacturing and metrology... how does one become more precise than input tools?
The ultimate reference for this is the Foundations of Mechanical Accuracy, which is available as a PDF[1], the print versions aren't cheap.
One example of precision is the Whitworth 3 plate method (invented in the 1830s), with this method, you can make surfaces that approach precision in the micron range, if necessary.
Leonardo Da Vinici's screw cutting lathe sketches show a machine that uses 2 lead screws to generate another, which averages the values of the other two screws, using careful rotations of threads, etc, it should be possible to use this method to work your way up to a uniform precise screw where each new generation is better than the ones before it.
[1] https://pearl-hifi.com/06_Lit_Archive/15_Mfrs_Publications/M...
The main way is by averaging error terms toward zero.
For example, Da Vinci's machine cuts the new screw blank in the centre of the carriage, which is driven by leadscrews on either side. The nuts would have likely been something like a wide strip of leather clamped over the screw thread, so there's enough compliance to average over a few pitches of thread, and the position of the cutter would be close to the average of the position set by each leadscrew thread.
Imagine the rough-cut screw threads have a pitch vs rotation angle described by p1(θ) and p2(θ). Running the machine then creates a new screw which is nearly a duplicate of the drive-screws in the machine but with pitch p3(θ) = (p1(θ) + p2(θ))/2. You can make two of these screws and swap them for the two leadscrews in the machine (it's built to be easy to do this). The random errors from a rough-cut screw gradually average out. But the cleverness doesn't end there. You then flip one of the screws backward end-to-end, so now you're averaging p3(θ) with p4(L-θ). You can also offset θ by any amount for either screw by offsetting the change-gear and re-clamping the carriage nut. Repeating these actions, you gradually can eliminate all systematic thread errors from the initial rough-cut screws and converge towards cutting a screw with nearly-constant pitch.
(It doesn't end there either; there's really a lot of flexibility with Da Vinci's design. Changing the gear ratio lets you create a fine-pitch thread from a coarse-pitch thread, or vise-versa, or cut a multi-start screw by rotating the blank 180 degrees or end-over-end).
The traditional way of making a multi-start screw is to use a dividing head and rotate the part an appropriate angle. If you're working with change gears, choosing a gear that's a multiple of the number of starts could provide alternate way to do this.
Oh yes, flipping the screw blank end-over-end would be an awkward way to do it unless you wanted to make a screw with two starts each having opposite errors. Once the screws have gotten good enough you wouldn't need to bother with things like that.
I should have noted that it's a built in dividing head that turns the part between starts.
Any time you let go of a part you're going to have to re-locate it before you start cutting again, and you will lose accuracy in the process.
Just some examples: take a string, don't bother to measure it, just any length between 1 and two meters or so would do. Take a pencil (or a piece of charcoal if you really want to go native) and a smooth branch. Stick the branch in the ground, tie the string around it so that it can slide with little friction and put the pencil in a loop of string on the other side. Now use this to create a circle. You started off with very rough elements not specifically sized for any purpose and ended up with a high precision representation of a mathematical concept.
Another: take a bunch of roughly cast metal balls. Put them on a sieve and let it vibrate until the balls have all passed through the holes in the sieve. Behold: metal spheres, so precise that you probably can't really measure the degree to which they are not spherical without resorting to instruments that you're not supposed to have in this scenario. Then sort by weight (which is a proxy for size). Now you can make ball bearings.
Yet another example: you can cut a lens for a telescope to within ridiculous precision using very primitive methods (https://www.instructables.com/Grind-and-Polish-a-DobsonianNe... ).
Put another way: it is always possible to increase your precision as long as you don't particularly care about absolutes or temperature effects.
Making sure you've seen this youtube channel, which is excellent: https://www.youtube.com/@machinethinking
I find this video of theirs the most relevant [0] where they go through how to start on a desert island and build a flat reference plate using the three plate method and the build up from there
[0] https://www.youtube.com/watch?v=gNRnrn5DE58
> I've let the folks at Wikipedia know.
What "folks at Wikipedia"? Can't you just edit the date yourself?
I could, but I just put a note on the "Talk" page, and let someone better at editing do it at some future date.
Have you ever thought about how alien lifeforms would probably invent screw cutting lathes too? The screw feels like such a "human thing", but what else would serve the rotational wedging purpose in this universe's elements and physics?
It would be funny to think of what might actually be a human thing. Like, our arms are quite weird, right? So potentially another intelligent species even on a rocky planet might not invent throwing spears, right? Even our close relatives, the chimpanzees, can’t use them well at all. Even fairly flighty animals seem to barely have the idea that a creature standing, like, tens of feet away from them might be “in range.”
I wonder what the trajectory (no pun intended) of the development of melee spears would be, without throwing spears.
The spear being a wildly popular a successful weapon for almost all of history, any changes to spears would, I guess, make a big difference.
Seems valid, I updated the tree to date the screw-cutting lathe to 1800
Super Easy Improvement: add thousands separator. It's easy to mistake 1,500,000 BC with 150,000 BC.
Also zoom in/out would be super useful!
Great idea though!
To add to this, anywhere you are showing numbers on a website, I'd recommend getting `Intl.NumberFormat` in use as soon as you can.
This 2D map is hard to explore since it's so sparse. I have to follow lines to find each thing, since it's 99% empty void. Is there a snap to next item hotkey? Am I just doing it wrong?
There is a "jump to nearest" button that appears if nothing is visible, but maybe it's too subtle? A hotkey would be a good idea too. I find that the best way to navigate is the minimap
Source code found here : https://github.com/etiennefd/hhr-tech-tree
https://github.com/etiennefd/hhr-tech-tree/blob/main/src/scr... this is kind of how I expected it. Honestly I would have done https://dumps.wikimedia.org/ and then parsed it.
Additionally I've always wanted institutions to be part of the timeline of technology. Corporations, Nation-states, Universities, Guilds, International Organizations - the ways people innovatively organize make things possible that otherwise wouldn't be.
The higgs boson experiments, for example wouldn't have been possible without the complex international institutions that orchestrated it. Manhattan project, Moon landing, the internet ... the iphone ...
This is truly amazing. I love it!
Found a minor nit — the entry for “shoe” is at 3500 BCE and links to the Wikipedia article for “sandal”. But the Wikipedia article for “shoe” [1] indicates that the earliest shoes (apparently not sandals) were worn around 7000-8000 BCE.
[1] https://en.m.wikipedia.org/wiki/Shoe
The links were mixed up (fixed now!) but I think the dates were correct; 3,500 BC is the date of the first leather shoe known, in Armenia.
I'd expect something things like Chinese Writing to be a big upstream dependency, but here it's a terminus. Detecting a western-bias in the sourcing.
Whether a node is a terminus (or root without predecessors) is basically ~never meaningful in the tree, it's almost always just missing data. Here it seems pretty clear I omitted a link from Chinese writing to woodblock printing. Fixed!
what's the technological difference between Chinese Writing and Western Writing?
why are they separate?
(Not an expert but) Chinese writing is hieroglyphic, wheres 'western writing' is phonetic. Western writing has a very small character set and is thus well-suited to a printing press, whereas hieroglyphics have thousands of characters (for thousands of concepts) but aren't fundamentally linked to the language like western characters are.
> Western writing has a very small character set and is thus well-suited to a printing press
This is vastly overstated. This was a widely popularized idea in the west but has largely been debunked by more recent scholarship that is less interested in demonstrating the superiority of the west.
Do you have links to relevant research?
Tom Mullaney at Stanford has a good book on chinese typewriters (this narrative is usually presented against both printing presses and typewriters) and is a good entrance into the intersection of chinese script and technology.
History research is typically published in books rather than papers, so it isn't content I can link to directly.
this is why Mesopotamian and Egyptian writing are upstream of so much.
upstream of what? Did you not look for mesopotamian or egyptian writing, which predate chinese writing by thousands of years. This sounds more like a chinese bias lol.
Such as?
I dunno man. Surely this is the sort of thing that it makes sense for a historian to do (they don't tend to like this sort of approach).
But they won't do it if they don't like this sort of approach! Which I think is wrong of them btw. But I don't think I've heard from historians much since this project went public.
Surely you'd understand how this sort of approach eliminates all contingency from the past. Especially given that tech trees are usually present in video games where they look forward rather than backward.
If you are doing a large project on the history of science (you have clearly put a ton of work into this) but you aren't communicating with historians of science that's concerning to me.
Related: The Universal Tech Tree - https://news.ycombinator.com/item?id=44161607
Thanks! Added to https://news.ycombinator.com/item?id=44829170 now.
If you like this then you will probably enjoy the book How to Invent Everything: A Survival Guide for the Stranded Time Traveler by cartoonist and computer-scientist Ryan North.
https://www.howtoinventeverything.com/
It's interesting that prior to the industrial revolution there are still some periods where it seems like innovations arrived relatively fast, and others where it was comparatively slow. E.g. a lot more entries are in the 500 BCE - 200 BCE period than the 200 - 500 range.
Although the idea of a "Dark Age" is mostly debunked these days, the slow unraveling of the Western Roman Empire led to a real and sustained change in material conditions. Notably, population density and urbanization both decreased, along with the labor specialization that accompanies them. I'd expect most 'inventions' to happen when and where people have the most hands on time to make them! (I can't really speak to Indian and Chinese civilizations, but they have also had integration and disintegration periods)
This is awesome. I worked on a 'conversational historical timeline generator' a little bit ago: https://timeline-of-everything.milst.dev/
I wonder if something similar could be added here where I say something like "what's the most important descendant of x" and it would bring me to that tech and give me a little explanation of why
This version of Sid Meier's Civilization would take ages to play.
This site is an absolute gem. Thank you.
Somewhat tangential, but can anyone explain or give any insight into how candles were invented? This has always puzzled me. The idea that if you put a piece of string in some wax, and light it on fire, the string will get hot enough to melt wax, and the molten wax will be sucked into it by capillary action, and burn, before the string itself burns, seems very counterintuitive and hard to come up with before a modern understanding of chemistry and physics. And yet, candles have apparently existed since 700 BC. How?
Pure speculation, but I would guess people first mixed wax and lamp oil in different ways to still get the burning effect of oil, with less of the cost of the oil, then added a wick to help light the oil/wax.
Then eventually that product morphed over time to the point where they realized the oil wasn't actually a necessary component
Good idea, it sounds plausible! But it still leaves open the question of how oil lamps were invented. How did someone figure out that a wick would be helpful?
Rope was around long before the wick. It seems conceivable that rope shavings or pieces or old rope were an easy way to start a fire.
This was then used with oil to make an even better fire starter or means of transferring fire. Eventually someone realises that a rope soaking in oil is easily lit and sustains a flame.
Before wicks, how do you burn oil? It's not easy to just ignite a bucket of lamp oil (putting aside what you might make the bucket out of). Probably you soak other fuel like wood or rags in the oil and burn the result. It's not a huge step from there to accidentally find out that you can make do with one piece of wood or cloth or string for a lot of oil, assuming you have something to put the oil in.
1. Someone dips a rope in flammable oil before lighting it, and sees that it's quite flammable. 2. Some other time, someone tries to use a rope dipped in flammable oil as a fuse to trigger the lighting of the oil once it burns back to the oil 3. They notice that the fuse keeps burning but doesn't burn back - in other words, the wicking effect 4. They shorten the rope and reshape the pot, and that's an oil lamp.
The first oil-lamps were basically a bowl with animal-fat and plant parts as wick. It's not hard to imagine how ten thousands years ago a hunting tribe could discover such a device by accident or on purpose.
Because the people from 700 BC were just as smart as we are. Observing the world around you is one of the best sources for ideas and there are predecessors to 'real' candles that must have fertilized the ground for the discovery. For instance, a stick dipped in rendered fat could serve as a torch. Not quite a candle but a significant step on the way there.
A way to zoom in and out would be very handy. On mobile it is pretty difficult to use as you need to scroll past vast empty spaces.
Pretty cool though and I learnt a thing or two. Thanks
It's funny that there are so many innovations right now the recent part of the chart just has to arbitrarily exclude an insane amount of stuff innovation that's happening.
No HIV vaccine. mRNA vaccine get's a single entry instead of vaccine per disease like prior vaccines. No battery stuff since 1985. Just amazing, fractal improvement is everywhere.
Great phrase - fractal improvement. It's kind of the idea of this book [0]
Even more cool: commercial progress trails tech. It takes a long time for companies to figure out how to turn a new idea or a cheaper input into a new product/industry, and then for related companies to grow into an economic ecosystem.
So one would expect to see some spectacular economics over the next couple of centuries.
[0] https://www.amazon.com/Abundance-Future-Better-Than-Think/dp...
Beautiful! I wonder if Jimmy Maher's heard about this; he wanted something like it for The Analog Antiquarian back ages ago before he kicked that off, as a way of reflecting the span of history in the structure of the index/TOC, but we never could figure out really how to get it to go anywhere we liked. It's a surprisingly tricky problem, and this is an impressive realization!
So, when's the flight to Alpha Centauri?
I'd prefer we only ever produced chariots and set tax to 90%.
This is cool, but I think the execution is off because there's so much empty space. I think it would work better if the nodes were much smaller and closer together so you can see more of the graph in one screen.
where is the zoom functionality??
Coming... soon...
Most comments here seems to be historical complaints. But I love this!)
Genuine question: how does it make sense that both bread and alcohol fermentation were invented before the domestication of any grain? Were people making bread (and beer!) with gathered seeds?
Yes, there are archeological remains of bread made from wild wheat and barley by hunter-gatherers in Jordan. It makes sense when you think about it, domestication would have happened only if the wild plant was useful, and if the plant was useful then it's likely people experimented with various ways of using it (like baking)
Wine is much easier than beer, because ripe fruit will ferment on it's own. Even wild animals get drunk sometimes; it's sometimes said that wine was discovered but beer was invented.
So alcohol without any grain is easy, but I don't know the answer for beer. On the other hand, why would you domesticate grain unless you already knew you could turn it into beer or bread?
Grain can be turned into porridge (so basically just soaked) and eaten, bread and beer is optional. In fact, IIRC classical-era armies would eat their grain as porridge instead of spending hours in camp baking it into bread. For obvious reasons.
Tobacco and alcohol are included, but there is no mention of other botanical or mycological developments outside food production. Those would be an interesting addition.
Such as?
https://en.wikipedia.org/wiki/Entheogenic_drugs_and_the_arch...
https://en.wikipedia.org/wiki/Psilocybin#History
https://en.wikipedia.org/wiki/Peyote
https://en.wikipedia.org/wiki/History_of_cannabis
https://en.wikipedia.org/wiki/Kava
Does anyone know which technology on this tree has the most descendents?
I vibe coded with gpt-5 and the source json (https://www.historicaltechtree.com/api/inventions) to get this list:
Top 10 inventions by number of direct descendants
1: High-vacuum tube — 13
2: Automobile — 12
3: Stored-program computer — 12
4: Voltaic pile — 11
5: High-pressure steam engine — 11
6: Glass blowing — 10
7: Papermaking — 10
8: Bipolar junction transistor — 10
9: Writing (Mesopotamia) — 9
10: MOSFET — 8
Top 10 by total descendants (direct + indirect)
1: Control of fire — 585
2: Charcoal — 444
3: Iron — 422
4: Iron smelting and wrought iron — 419
5: Ceramic — 404
6: Pottery — 402
7: Induction coil — 389
8: Raft — 365
9: Boat — 363
10: Alcohol fermentation — 353
Top 10 by total ancestors (direct + indirect)
1: Robotaxi — 253
2: Moon landing — 242
3: Space telescope — 238
4: Lidar — 236
5: Satellite television — 231
6: Space station — 228
7: Stealth aircraft — 228
8: Reusable spacecraft — 224
9: Satellite navigation system — 224
10: Communications satellite — 224
Highly recommend the Dr. Stone anime if you're interested in a story with the premise of starting civilization from scratch but armed with the sum total of modern human knowledge about science and engineering.
Also, if you want even further back precedent for this kind of plot device, I highly recommend reading a Connecticut Yankee in King Arthur's court by Mark Twain.
https://en.wikipedia.org/wiki/A_Connecticut_Yankee_in_King_A...
I second this. It's the only show I've seen making a semi-realistic attempt at this (ignoring the absurdity of the initial petrification in the first place and Dr. Stone having superhuman knowledge of all human inventions)
I'd also recommend the Destiny's Crucible series - the basic premise is that a chemist from our world is transported to another planet of humans at a much lower technological level, and some moderately standard isekai hijinks ensue.
I read five of the books, and really enjoyed them; if you like the "competence porn" genre of novels, this is a pretty good one.
> "competence porn"
See... now, I love that type of show/comic/book/etc. And now that I have a name for it, I want to search for more. But I very much do _not_ want to search for that term. Lol
I think a similar genre is "humanity fuck yeah" - HFY - so you can search for that as well.
I watch this with my daughter and we love it. I love shows with "narration", talking about the context/details of things, and Dr Stone really nails that (I know the main character isn't really a narrator.. but it accomplishes the same thing).
I'd also recommend the "How to Make Everything" YouTube channel.
A tech tree without metallurgy roots is like building a spaceship with no screwdriver—looks cool, but it’s gonna fall apart fast.
Cool concept. I’d love a vertical version for mobile.
Looking forward to the new Civilization mod that uses this.
Reminds me of the tech tree featured in the game Civilization. Pretty cool stuff
Cool. So when was the nail cutter/clipper invented?
Very cool! Will explore it a bit :)
No fire, and no knot. Hmmm...
Yes there is: "control of fire". No knots, but ropes around 50000 BCE.
"Control of Fire" is right under 1000000 BC
Ok, I'm just bad at search.
Very cool, cheers!
have always wanted something like this! awesome!!
Its a great start! Bound to have bias and blindspots. It would be cool to run an agent that could incrementally enrich this knowledge graph. Take some modern day technologies and backtrace the components and their development.
And a related page, in the other direction: https://www.futuretimeline.net/
In what sense related?
This is really cool but hard to view well on a PC. I'd love to have a simplified version of this on a big A2 poster.
Pretty cool. Makes me think if we're overdue for another 1960s era tech boom?
1760000 BC: StoreTool 3. This is our greatest model yet. You are going to love it.
These paradox games are getting out of control