One of the biggest intuitive mysteries to me is how humanity took so long to do anything.

Humans have been ‘behaviorally modern’ for about 50 thousand years. And apparently didn’t invent, for instance:

This kind of thing seems really weird introspectively, because it is hard to imagine going a whole lifetime in the wilderness without wanting something like rope, or going a whole day wanting something like rope without figuring out how to make something like rope. Yet apparently people went for about a thousand lifetimes without that happening.

Some possible explanations:

  1. Inventions are usually more ingenious than they seem. LiveScience argues that it took so long to invent the wheel because “The tricky thing about the wheel is not conceiving of a cylinder rolling on its edge. It’s figuring out how to connect a stable, stationary platform to that cylinder.” I feel like that would explain why it took a month rather than a day. But a couple of thousand lifetimes?
  2. Knowing what you are looking for is everything. If you sat a person down and said, “look, how do you attach a stationary platform to a rolling thing?” they could figure it out within a few hours, but if you just give them the world, they don’t think about whether a stationary platform attached to a rolling thing would be useful, so “how do you attach a stationary platform to a rolling thing” doesn’t come up as a salient question for a couple of thousand lifetimes.
  3. Having concepts in general is a big deal, and being an early human who had never heard of any invention was a bit like being me when I’m half asleep.
  4. Everything is always mysteriously a thousand times harder than you might think. Consider writing a blog post. Why haven’t I written a blog post in a month?
  5. Others?

New Comment
17 comments, sorted by Click to highlight new comments since:

My cousin suggests this is false because, while things can be discovered many times, it has to really take off and get wide adoption for us to find it historically.

This seems plausible for fire; it took ages before wide adoption of fire in humans, but I read a recent article that some present day bonobos discovered fire. If it can happen in my life time, this suggests it would happen many times in history, it just didn’t take off.

I think the same goes for present day inventions. The fact that Yudkowsky invented a way to cure his wife's SAD by putting up enough lamps doesn't mean that the invention immeditately spreads to everybody.

In The legend of healthcare, Michael Vassar makes the case that even through medical books talk about the invention of using mirrors for treating pantom limb pain, the treatment usually isn't used and doctors rather give painkillers.

It would be quite easy for our civilisations to forget the invention.

It's widely believed that humans basically lived in <200 person tribes that didn't interact with each other too much before agriculture, so one might wonder how anything got any amount of adoption.

And only about 100 years ago, humanity essentially forgot the cure for scurvy: http://idlewords.com/2010/03/scott_and_scurvy.htm

It might be profitable to take a recent breakthrough and see why it didn't happen before.

When I did this I found

a) Often an invention requires a lot of prerequisites. Useful wheels seem to require some strong metal as a core component. Also they need open country and smooth hard surfaces to run on. Deep learning required fast computers with a lot of memory.

b) Some actually important inventions are not seen as useful. The Romans had little use for labor saving devices because they had slaves.

c) There are many ways to do something that are obvious, plausible, simple, and wrong. E.g. Edison's famous comment about the light bulb.

d) Some inventions go against received wisdom. Stretching the definition of invention little, Kepler discovered of the elliptical orbits of the planets and pubished the results, stating that it could not be correct because as we all know the true orbits must be based on circless.

As an exercise, take a current unsolved problem and try to solve it. Consider friendly AI.

What are you up against? <This is a stupid problem and does not require a solution> <We will not understand AI until we understand consciousness> <True intelligence in the human brain comes from quantum magic> <It cannot be solved> <Just turn the thing off if it becomes a problem> Etc.

Or better batteries (100X better batteries would solve so many problems).

Your point c definitely rings true to me. An answer often seems simple in hindsight, but that an answer is simple doesn’t mean it’s simple to find. There are often many simple answers and the vast majority of them useless.

I am an inventor. I find the process of inventing is an exercise in multiple streams of coexisting thoughts. For instance, first defining the problem is paramount. But after that I need to both understand existing solutions to see how they could be improved but also erase all existing solutions and imagine what solutions I can from scratch. Also, I need to imagine desired solutions with producible solutions. I need to pursue solutions but also periodically abandon them all and go back to square zero. I am trying to describe having multiple thought trains which, if not symultaneous, alternate in as complete a way as possible (where one disappears when the contrasting one is operating).

Whether a solution is an invention depends on if the solution is new or not. I may come up with something new to me which I discover someone else already came up with, thus it isn't offcially an invention. That's why I may use the word "solution" instead of "invention".

Then there is the wrinkle of whether a solution can be manufactured, whether as a one-off or in quantity. And the detail of whether materials exist to make it. And then the need to both take into account existing materials and also to try to think of "what if" there weren't that limitation. So often one can be self-limited by using known materials or processes. By thinking "There must be a way to…" or "Someone must know how to…" or "How hard can it be to…" can lead to dead ends or new solutions.

An interesting thing is using CAD to draw and a 3D printer to produce something. It is marvelous because I get to see it on-screen and to produce it effortlessly. I can make a change and then print it out without spending hours cutting, gluing and sanding. But at the same time I am limiting myself with what drawing skills I have (or the capabilities of the program) and by what the printer can produce. And then there is the maddening thing of also being haunted by what can be manufactured (3D printing and injection molding have distinctly different capabilities and limitations) which can limit me prematurely if I let it.

I sometimes try to transport myself back in time and try to imagine what I would have done when stone, bone and such were the materials to hand. As someone else noted, why work on an axeled wheel when no roads were available? A travois may actually be a better solution on that day. Or what would I imagine to be a problem that needed to be solved? What would improve the life I actually faced each day? How much of my life would be consumed with just learing how to do stuff that already existed? And how many inventions were the mavels of their day but then ceased to be preferred?

Oh, which leads to the topic of un-learning wrong information one has picked up. Or value judgements on what is good or bad.

And then there is the dual processes of imagining how something might work and actually making a prototype and seeing how it actually functions.

And then there is the aspect of whether something is truly new or is an improvement on an existing product. New products are difficult to introduce because they require educating users. Improvements may be too expensive for companies to justify. Either can canabalize sales of existing products yielding no net gain. So all the solutions cause problems for companies at the same time as they offer tantalizing new products.

And then there is the limitation of using words to describe solutions. Or of understanding of machines (basic ones like ramp, screw and so on). Knowledge is a blessing and a curse. A blessing in order to combine wjhat one knows, but a curse because it can fence one's thoughts in. A great example I used to use was how easy it was to spot an advertisment which had been created with Word…lots of text with boxes and so forth (generally looked terrible). Someone who used Quark Xpress or Illustrator would have a more flexible canvas to work on (and could produce more artistic or exotic stuff). But either way the tools and knowledge allowed for marvelous new things and limited one at the same time.

My point is how many counterposed or mutually exclusive thought trains are involved.

Oh and to answer your first wondering: yes, inventions are more ingenious than folks may suppose. That's why I am such an impressive guy. ;-)

In addition to what others said, things get rediscovered many times before the discovery takes hold. Think of it like a beneficial DNA mutation. The circumstances have to be just right for it to survive. In retrospect we find a lot of discoveries that were forgotten for years, decades or centuries. And earlier in human history those would not leave any artifacts for us to find.

Regarding the wheel

I don't find it mysterious at all why wheels took "so long". I'd expect the wheel to be conceptually discovered much earlier then it's first usage, because to be actually useful in the real world a wheel requires roads or vast flat plains to outperform using a backpack, let alone simply loading the animal with sacks in case your civilization figured out animal husbandry of at least one animal useful for transporting.

Regarding printing

I was always a lot more surprised by how long the Gutenberg-type printing press (i.e. movable type press) took to invent and take off. According to wikipedia the earliest ones were invented around 1000-1200 B.C. in Asia https://en.wikipedia.org/wiki/Printing_press

Compared to other feats of engineering of the time you'd think arranging single letters on a slab and printing a few hundred pages rather than having monks do all the painstaking handwriting must have been an utterly obvious invention (though there may be a prestige factor in play here, since obviously handmade writing with illustrations is much more beautiful and would be preferred by people who could read, which were to a large extent also the people who'd be able to afford it in the first place).

However, upon looking into it, the invention of the "hand mould" and its manufacturing precision seemed to be of much importance for the technology to take off. Precise manufacturing of type seems to be of huge importance, because if some of your letters are just a tiny bit shorter or longer than the others it won't work at all and you'll get missing letters or otherwise low-quality print (e.g. if the letters are not at the same height). Since printing quality before Gutenberg was bad, people who could actually read presumably wouldn't have wanted to waste their time with this low-status quality nonsense anyway. Unfortunately there is no wiki article in English, but check out these pictures to get an idea how the first hand moulds actually looked like: https://www.buch-kunst-papier.de/drucken/raritaeten/handgieinstrument.php

Basically people molded the type pieces directly in their hand inside a small device that could be separated into two halves. The type letter itself was stenciled into blocks of metal that would then fit inside the handheld mould device perfectly, achieving the needed standardized precision. Check out the pictures in this slider to get the idea: http://www.druckkunst-museum.de/de/schriftgiesserei.html

So it's not that no one has conceived of it or tried it before, it's just that the quality was too bad and no one figured out how to make perfectly standardized metal typesets. So Gutenbergs achievement wasn't really so much the obvious basic idea of movable type, his real achievement was actually the invention of the hand mould, which is the much less obvious part of making the printing press actually work with enough efficiency to be worth the trouble.

Also, see https://en.wikipedia.org/wiki/Johannes_Gutenberg#Printing_method_with_moveable_type

Thus, they speculated that "the decisive factor for the birth of typography", the use of reusable moulds for casting type, was a more progressive process than was previously thought.[32] They suggested that the additional step of using the punch to create a mould that could be reused many times was not taken until twenty years later, in the 1470s. Others have not accepted some or all of their suggestions, and have interpreted the evidence in other ways, and the truth of the matter remains uncertain.[33]

Gutenberg didn't get rich through his Bible printing business. He didn't make enough money with his printing business to pay his creditors and lost his business as a result.

For the printing press to be more economical than hand written books you actually have to sell many copies of your book. Books used to be written on parchment and that's really expensive and would likely have been to expensive for making the printing press work.

Paper that's much cheaper than parchment was likely a necessity for the printing press to have commercial success. According to Wikipedia the first paper mill in Germany was build in Mainz in 1320. By random Mainz happens to be the city in which Gutenberg was born.

Gutenberg burrowed money to employ 20 people over two years to print 180 copies of his bible. If you are used to books being very expensive the prospect of it being responable to print 180 copies of a single book might seem far fetched and as a result few people would start such a project.

The fact that he actually could borrow money to run a business with 20 people is notable because that means that certain economics have to be in place that weren't at many times in history.

Update: I did some further research and it's possible that the paper mill in Mainz didn't exist and the first German paper mill was created in 1390 in Nürnberg. If that's true that disturbs the narrative of the story a bit, but it doesn't change the fact that paper availability was essential for Gutenberg.

Books used to be written on parchment and that's really expensive

I bet people don't quite realize HOW expensive it is. Even today, a single piece of parchment the size of a piece of notebook paper is going to run you about 50 dollars. Now make a book of them.

Do you think you could make a rope in a single day when you would be in the wilderness with only an obsidian knife as your tool in a single day?Even knowing how ropes work I would estimate that it would take me more than a single day to come up with the necessary knowledge to create rope.

Especially without any knowledge of rope I think it would take me a lot longer.

I'm curious about your item three.

Nobody told early humans to invent things. They just had to end up doing it. That's also true for crows and other primates. If you were a crow, how would you find and use a tool? (Warning: I'm trying to work toward a plausible story in the following. There are probably lots of wrong implications about animals.)

Clavicus the crow flew straight over the field to a new tree. It had seen the setting sun and knew that meant it was time to return home. Every time clavicus went to a tree, it thought for a moment about where it had bee n last, so it would find the place with the tastiest and newest worms. All of this happened in a flash of its mind, which no faculty of its own was aware of.

Clavicus thought about little when it slept, and when it woke up, it thought about worms in the ground. Each day, it went down to the ground and picked at it. It had learned to identify many promising signs of worms, even above the obvious little burrows. Parts of the field covered in grass were more promising than barren patches of dirt. The latter were easier to look at, but could be scanned quickly during flight for burrows, saving time to look in denser areas.

Clavicus went flying out on Yellow-Grass Prairie With Triangle Of Trees In The Center the next morning. As it searched below Tree In The Triangle Closest To The Place Where It Made Its Nest, Clavicus's beak bumped against a stick, which upset the dirt below it, leading to a large upset of the dirt, a sign that Clavicus at this point recognized as being great indeed.

It was bound to happen, once every million crows.

Clavicus tried the movement again and soon was able to do it pretty quickly and with little error, as inefficient ways of moving the stick got pruned away from its thinking.

This was only bound to happen once every billion.

The next day, Clavicus got eaten by a cat. Some other crows eventually used tools, and their relatives saw them and imitated them. Soon, tool use was common enough among crows that it persisted.

Animals that couldn't imitate each other weren't so lucky as the crows.

Reality has a surprising amount of detail (there was a post about this explaining it at the example of contructing a simple wooden ladder which I can't find, but I bet there are a lot comparable descriptions out there). Or take a candle. I guess you have used one recently. Looks pretty simple, right? Just use some wax and a wick. Turns out that people have used candles since ages. They were frequently used in rome for example. But the easy to use candles of our time are pretty recent. Recent as in last century. Before that

  • they didn't have wicks that burned themselves away and you had to cut them all the time
  • there was no good wax. Most candles were made of fat with lots of residue that stank and smoked. Bees wax was much better but harder to get

To fix these things you need much better raw materials and production processes...

See this article about candle history (German, but I guess Google translate is good enough).

And you can look at any kind of thing we take for granted and it is basically not posible to grasp all of it. The classical example is I, Pencil: My Family Tree as told to Leonard E. Read Most things depend on the presence of a whole environment - and take part in bringing it about. You could see it as a co-evolution of lots of inventions. Something just hinted at in the comment about roads needed for wheels (and actually you benefit from having wheels when building roads...).

I think this is one of the main overlooked points when talking about the possibility of space travel, esp. interstellar one. Even if you assume AIs. But let's not. As mentioned in another comment we not really know what kind of coordination problems it comes with. Scaling isn't automatic. Just look at Moore's law. Sure we continue to scale, but we pile technology on technology on technology to do so, And we can't just invent the last one. And neither can a future AI. You need the whole stack (OK granted, you might be able to simplify, but still). And it will keep growing and might become inherently unmanageable. Remember: The price of a chip factories also continues to grow and that might be the limiting factor. See e.g. McKinsey on Semiconductors 2013.

You were referring to: http://johnsalvatier.org/blog/2017/reality-has-a-surprising-amount-of-detail

One of the biggest leaps I made in trying to understand innovation was:

A. Realizing that new technologies and ideas evolve from misunderstood to understood, and only in the process become truly world changing.

B. Realizing that many (most?) new technologies and ideas rely on old ideas and technologies being understood to the extent that they are commodities or utilites.

Note that A is actually dependent on the speed of communication, so our current intuitions about how fast A happens are orders of magnitude off base.

Two of my favorite mental models to help make sense of how this works:

The Carlotta Perez Framework

Wardley Mapping

Moved to frontpage.