Can advanced technologies be used without understanding them?

[Ramza00]

Not enough of a sci fi geek to have actually read it, but to my understanding the Foundation books by Asimov had a "Church of Science" where they did not really taught science to most people, but the high priests knew how science operated, yet the lower priests and the public saw it as magic. At first these magical items required you to set up the social structure of the religion on your planet, but later on this policy is changed and the items could be freely traded.

[SlyGuyMcFly]

Can a civilization theoretically utilize and replicate advanced technologies in their day-to-day life without an intimate understanding of the scientific principles which allow the technologies to function?

Sure. They don't need to know why it works. Or even how it works. Just how to make it work. How to build it and what to build it from, basically. The civ in question may have no idea the precursor ships syphon energy from the zero-point whatever to power the core. They just need to study said core long enough to figure out the Ancient Ones put a space-mythril filigree around the main post-transuranic crystal array in such-and-such pattern because otherwise the engine collapses into a black hole. Or makes an annoying clunking sound that goes away for five minutes if you thump it just right but it's a real pain. Depending on the setting's tone.
Repeat for a hundred different technologies and there you go. They can build it, use it, sometimes fix it and never improve it.

This does require the archaeologist civ to reach sorta-close to precursor tech in some areas at least. They have to be able to produce the raw materials such as exotic alloys by themselves. They'd also need to be able to build the tools needed - if your parts need to be laser-etched and you ain't got lasers, well ain't that a kick in the pants. However, when you know you need something that can do what a laser does, inventing lasers is a lot easier. So maybe half their tech is stuff they figured out from looking at pre-tech, the other half is pre-tech they just duplicate using the first half.

A way of making it more plausible is that they found a factory, or derelict foundry ship, or something along those lines. They could have duplicated the production process without understanding the principles via trial and error.

[Tyndmyr]

*looks at the internet*

Yes.

[wumpus]

Heavier-than-air flight is a lousy example, since we've understood how it works for significantly longer than we've been doing it.

Which explains why the Wright Brothers famously invented the wind tunnel and promptly threw it out because the knowledge of the day was sufficient. Of course it really wasn't and they had to do considerable experiments to understand heavier than air flight (and more importantly learn the importance of banking to turn).

Of course for a larger picture, you are correct. With sufficient power you can make a brick fly, and about 10 years after the Wright Flyer flew, you didn't even need pre-flight information to get your plane off the ground (all though it wouldn't be all that efficient). Often the issue isn't "do you understand" but "how close is your understanding to the underlying reality".

One thing that hasn't been brought up is that typically in modern tech *nobody* knows how it all works. Or more specifically, no one individual knows how it all works. Technology tends to work like a software stack or an assembly line where each engineer takes the design from the last guy and adds his knowledge to the "technology" and passes it to the next cubicle. Sometimes this might be ordering a particular steel (which requires knowledge from everything from geology to find the ores, mining, smelting, and finally metalurgy to create the right steel) and then shaping said part via CNC. The whole part then protects a laser designed by a physicist and built by people with the dexterity the physicist never had (in a way he never thought of). The laser is then controlled by a programer who doesn't even know that it is a laser (he wrote it assuming cat6) and is finally used to run the latest pay2win mobile game.

But almost nobody has to know how the other guy's tech works. They might not even be aware that they are working "together". Its hard to say when tech got this way, but in 1556 Georgius Agricola wrote "De Re Metalica" a book that basically detailed all 16th century knowledge on mining an metalurgy, so presumably people could still no *everything* about a technology. About this time Guetenburg was producing [western] movable type, an invention that required tons of non-related details (needed paper, the right type of ink, specific designs to get type sufficiently regular. The "printing press" was the easiest bit) that could presumably be invented by a team that doesn't understand each other (previous "great inventions" were pretty much a single idea). This isn't to say that everyone knew how to build and use Egyptian canals, just that we know when someone *could* understand all tech, and when it was more likely they couldn't

[Bohandas]

Can a civilization theoretically utilize and replicate advanced technologies in their day-to-day life without an intimate understanding of the scientific principles which allow the technologies to function?

I've put some thought into a similar question of whether people would be able to reverse an MP3 player that was sent back in time. I think Nicola Tesla might have a shot at it. Maybe Franklin if he had a whole case of them to experiment on, but proba ly not anyone earlier than that.

[halfeye]

I've put some thought into a similar question of whether people would be able to reverse an MP3 player that was sent back in time. I think Nicola Tesla might have a shot at it. Maybe Franklin if he had a whole case of them to experiment on, but proba ly not anyone earlier than that.

There are chips in those, so I'm pretty sure the answer is no, Franklin and Tesla wouldn't have a clue. There are patterns on chips, which can be read if you know what you're looking at, but most of the parts are transistors, and transistors weren't made until after WW2.

Once you know what a transistor does, making a duplicate of a chip on a much larger scale is possible, when you have the etching (or whatever, I'm not exactly sure myself) process then you can copy chips, but the least you need is transistors, without knowledge of what transistors do, you can't guess what the circuits that contain them do.

There were valves:

https://en.wikipedia.org/wiki/Vacuum_tube

They were most common between the wars, they were generally much more complicated than transistors, but much less complicated than chips, you couldn't extrapolate transistors from valves, they're not similar enough.

[factotum]

There are chips in those, so I'm pretty sure the answer is no, Franklin and Tesla wouldn't have a clue. There are patterns on chips, which can be read if you know what you're looking at, but most of the parts are transistors, and transistors weren't made until after WW2.

More to the point, they probably wouldn't have a microscope with sufficient resolution to actually see the features on the silicon, even if they successfully managed to remove it from the packaging without damaging it--the electron microscope wasn't invented until 1933, and while Tesla could conceivably have made use of such a thing he was already 76 at the time and might not have been quite as sharp as in his younger days!

[Lord Joeltion]

More to the point, they probably wouldn't have a microscope with sufficient resolution to actually see the features on the silicon, even if they successfully managed to remove it from the packaging without damaging it--the electron microscope wasn't invented until 1933, and while Tesla could conceivably have made use of such a thing he was already 76 at the time and might not have been quite as sharp as in his younger days!

I think the biggest issue is that, even if they replicated the device by blind chance; they would still be unable to use it, because they have no data or method to transfer any kind of information the MP3 needs in order to function. Which means that even with schematics of transistors, microchips et al; without the concept of programming, the device is useless, making the whole task a pointless endeavour.

I think problem isn't whether you understand how it actually works or why; but whether you have at least some rough knowledge of the concepts by which the device works and its intended use. Otherwise, any kind of research will wildly deviate from the correct answer.

Give a toddler an iphone and never tell him the password to use it: he will forever believe it's a nice lantern/mirror rock. He will only recognize it as a communication device/miniature computer once he acquires the concepts of "communication" and "data entry", not before*. Technology is (more often than not) intrinsically related to its use.

*And even then, he will believe it's powered by "magic" until he discover the meaning of "electronics".

[Bohandas]

I was thinking an MP3 player with some songs preloaded on it

[wumpus]

I was thinking an MP3 player with some songs preloaded on it

[TL;DR] Once engineers understand a technology, they keep working on it till it isn't remotely understandable (if it works it isn't optimized enough).[/TL;DR]

Still impossible. It is entirely possible that it wouldn't directly help produce MP3s if sent back to 1980.

You assume that Tesla is shown how to operate it. He knows that there is music stored inside and can be convinced to play it. He also carefully measures the battery and knows not to use voltages higher than it produces (still likely to fry the chip, but we'll ignore that).

First, you are wildly underestimating the complexity of *any* modern system on a chip. Once a tech matures, the original simple idea is lost and it becomes covered with tiny evolutionary improvements which completely obscure the original intent. A much better idea would be to send an early 1970s calculator. He might have a chance at that.

Consider explaining a car to Franklin. If you show him a Prius (or any 21st century car) and you will never get anywhere. Every single part is likely too complicated to begin to explain. Start with a Model T and he would understand it in a week (considering the toolkit needed to repair a VW beetle, that would likely work as well).

Start with the issue of pulling the data off the flash. It probably comes out at between 0-1.x V, but vacuum tubes love wildly higher voltage (although obviously they can amplify small signals as that is the point of radio), and he has to carefully pull out the data. If he manages this, he gets nothing that looks anything like music. MP3 codes thing as huffman encoded heavily reduced modified discrete cosine transforms. Trying to decode huffman encoding might be remotely possible (telegraph users were getting the idea) if he had some idea of what he is looking for, but the idea that it would be a few DCT coefficients (with windowing overlap, something that isn't remotely expected until you try using incomplete coefficient values of a DCT). So the data doesn't make any sense, and I wouldn't expect Turing himself to break the code.

The chip is even worse. Assuming the chip is sufficiently old/small (and uses a .28 process, more modern high volume chips are unlikely to be all that visible under a microscope) the shear *amount* of transistors is far too large to comprehend (I'm assuming tens or hundreds of millions). Much of them used in ways that make no sense to anyone who hasn't laid out such a chip (and thanks to Verilog and similar, most aren't laid out by humans).

A 1970s calculator might be remotely possible, although I'd highly recommend more than one of them (you can't find the "schematic" of a chip without grinding it down and making it unusable, and then you need to check your assumptions against a working model). You would still want a strong microscope and an oscilliscope, but you might actually be able to use both. The chips themselves were likely hand laid out with strips of mylar to hand-craft the mask (I think, they may have produced the mask with the mylar. But it was directly used to create chips or tooling) so it is reasonable to assume that somebody understood it when they laid it out (they were made by individuals or small teams, unlike modern MP3 players) and therefore someone can understand it by reverse engineering. Also, the operation of a calculator is pretty straightforward, at least once you slow the clock down enough to understand how it controls the LED displays (that can be somewhat non-intuitive), fortunately the clock is external and likely familiar to Tesla.

[Yora]

An interesting example of a not understood technology is anesthesia. You give a patient a drug, it goes to the brain, [and then something magical happens], and he's out. We have the chemical structures of the drugs which allows us to make them and we know how to use them. But we don't really know what they do.

[Lord Torath]

Interestingly, aside from improvements in battery technology (and the proliferation of electronic features), electric cars in the early 1900s were pretty similar to those today. They had patents for regenerative breaking, hybrid gas-electric cars, and Edison even patented an electric car motor that can be converted to a generator to charge the car batteries.

[Knaight]

Interestingly, aside from improvements in battery technology (and the proliferation of electronic features), electric cars in the early 1900s were pretty similar to those today. They had patents for regenerative breaking, hybrid gas-electric cars, and Edison even patented an electric car motor that can be converted to a generator to charge the car batteries.

There's also some pretty dramatic changes in terms of shape and material - and that motor/generator is an incredibly basic device once you already understand how to make a motor and a generator.

[factotum]

It's a matter of how far back you go, isn't it? Yes, an engineer from 100 years ago might be able to look at a modern electric car and mostly be able to figure out how it works, apart from the computerised control system. Go back 300 years, though, before even the simplest electric motors were created, and things get a great deal harder.

[Gnoman]

There are chips in those, so I'm pretty sure the answer is no, Franklin and Tesla wouldn't have a clue. There are patterns on chips, which can be read if you know what you're looking at, but most of the parts are transistors, and transistors weren't made until after WW2.

Once you know what a transistor does, making a duplicate of a chip on a much larger scale is possible, when you have the etching (or whatever, I'm not exactly sure myself) process then you can copy chips, but the least you need is transistors, without knowledge of what transistors do, you can't guess what the circuits that contain them do.

There were valves:

https://en.wikipedia.org/wiki/Vacuum_tube

They were most common between the wars, they were generally much more complicated than transistors, but much less complicated than chips, you couldn't extrapolate transistors from valves, they're not similar enough.

One caveat here.

Any device built entirely with discrete transistors can be duplicated with vacuum tubes, assuming you can figure out the exact type of transistor. This is because each type of transistor is a 1:1 solid-state analog to the types of vacuum tubes in use when they were developed. Such devices will work much less efficiently and draw more power due to the nature of tubes, but it is possible.

Once you get into ICs or compound transistors, this goes out the window, and virtually everything uses at least some of these nowadays.

[Lord Joeltion]

It's a matter of how far back you go, isn't it? Yes, an engineer from 100 years ago might be able to look at a modern electric car and mostly be able to figure out how it works, apart from the computerised control system. Go back 300 years, though, before even the simplest electric motors were created, and things get a great deal harder.

I don't think it's a matter of "time", but the technology you have at your disposal. Suppose a world with literally zero ferromagnetic materials, and magnets will forever be a mystery for the people that live there.

The same way, you need a certain basic technology in order to understand more advanced technology, or at least try to reverse-engineer it. Paraphrasing Newton: "you need to stand on the shoulder of giants". IMO, an extremely specialized society would have a much harder time trying to figure out a relatively "simple" device* (compared to their own tech level) than a less advanced society which is, on the other hand, more of a generalist type**. That is because technology is always progressive, and most of the time, there are no shortcuts or a path around a specific problematic. You still need the necessary basic tools to even begin research. There are various examples of fields of theoretical science which we had to wait decades to properly research, because we still lacked the basic tools for the task. I think the same may apply to advanced technology.

Of course, for relatively simple technology the rule does not apply. Cars are much more close in complexity to steamboats than they are to computers, even when they are closer to the latter in terms of time. That's why cars are way easier to reverse engineer, even if you go centuries back. For computers (and specifically, microchips or software) not so much.

*That is, a device completely unrelated to their ridiculous specialized technology, of course.
**And if they have a "less advanced" type of device with a similar purpose.

[wumpus]

Interestingly, aside from improvements in battery technology (and the proliferation of electronic features), electric cars in the early 1900s were pretty similar to those today. They had patents for regenerative breaking, hybrid gas-electric cars, and Edison even patented an electric car motor that can be converted to a generator to charge the car batteries.

There are also some profound differences. Teslas use induction motors, which while commonly used in the 1900s were most certainly not used in most cars (non-battery hybrids might be an exception). Induction motors were called "AC motors" and batteries only produce DC. You need to create your own AC (via a switching inverter/power supply) which wasn't viable until at least the 1980s (and probably too weak for cars until much later, like the EV1). Similar tricks are needed to get anything close to efficiency out of an electric car, and it is no surprise that gasoline was preferred from 1910?-2020 by most drivers (even hybrid cars often aren't quite there, but I have confidence they will begin to replace purely gas cars).

Pretty much all the technologies used in modern cars were invented by the 1920s (VTEC and similar might be an exception, but don't be too sure), but inventing the most primitive example and being able to build it robust enough to last the life of a car and efficient enough to warrant inclusion are far, far, more difficult than the "proof of concept" we saw during the 1920s (and earlier for electric cars).

This is astonishingly true for nearly all industries. Pretty much all possible improvements are obvious given enough eyes, but not all such improvements are economically viable (although given time they will be). Back in the late 1990s there was a bit of a shakeup in wireless communications: somebody invented a means of correcting errors at near the theoretical limit (Shannon Limit) of such things (the turbo codes). It was so unexpected that some were calling it a fraud. Of course, it was patented (if people who should know better are calling it a fraud, you can be fairly sure it passes the "nonobvious" test) so people were looking for ways around it. Way back when, the guy who literally wrote the book [well, preferred textbook] on error correcting codes, Gallager, had written a doctoral dissertation on a means of getting close to the Shannon Limit. Since it would require an unfeasible amount of cycles to complete the calculation, he didn't include it in his book. However, by the late 1990s an "unfeasible amount of calculations" were orders of magnitude higher and a bit above what was needed for Gallager's system, known as LDPC. It had been sitting under our noses the whole time.

[Vykryl]

Look into the Warhammer 40k universe. Man can has computers and replicators that read and build everything from STC files (standard template construct). Most vehicles are modular and can have components swapped back and forth. Very little is understood about how anything works. They push the appropriate button on the consul and out comes the part. They do know how to read repair manules to figure out how to change the components. If their machinery wasn't self diagnosing they wouldn't have a clue on what was wrong with it when it stops working.

Edit: BattleTech also has a similar situation with their jump ships and hyper pulse generators (galactic communication system). They can follow schematics to make components and change them out but have no real idea how anything works.

Care of advanced tech in both realms borders on a religious affair.

[factotum]

Look into the Warhammer 40k universe. Man can has computers and replicators that read and build everything from STC files (standard template construct). Most vehicles are modular and can have components swapped back and forth. Very little is understood about how anything works. They push the appropriate button on the consul and out comes the part.

But then you potentially get into the situation of the E.M. Forster short story "The Machine Stops", where humanity are entirely reliant on a giant machine to meet all their needs and are thus in a great deal of trouble when it breaks down and no-one knows how to fix it. (I know you said WH40K machines are self-diagnosing, but what happens if the self-repair bit breaks down?).

[Nourjan]

I'll chime in a 'probably yes'. I recall a thread a while ago about what if modern firearms were teleported back hundreds of years. I think there was some talk that the lack of tools to replicate the precision of modern parts and gun barrels would limit what could be replicated, but it would certainly allow use of some modern gear and replicating what could be replicated.

Could you link me to that thread or give me its title so I can look for it myself ,please?

[bulbaquil]

- While the device itself is still operational, of course it can be used (assuming they can figure out, through trial and error, which button does what etc...). By extension, if a device is self-repairing, it can be used indefinitely (at least until the self-repair mechanisms break down). Cities were using Roman aqueducts long after the engineering base to understand how they worked fell with Rome.

- The more mechanical the device (as opposed to electrical), the easier it is going to be to reverse-engineer, since a tinkerer can open it up and actually look at all the moving parts. To reverse-engineer an electric device, you'll probably need to have some cultural understanding of electricity first.

- If the device requires fuel, it will depend on how recognizable the fuel is (e.g. coal is straightforward; gunpowder or gasoline might require some chemistry knowledge) and how reproducible the fuel is with readily-available technology. If you can't produce gasoline or engineer the car's mechanism to run on something other than gasoline, you're getting whatever's left in the tank and that's it.

So if a culture at, say, Bronze Age tech levels or higher encountered a steam engine, they could probably figure out how it works and how to build another, even if they wouldn't know the thermodynamic principles that underline why it works. To build and use a microchip, on the other hand, you'd probably have to be fairly close to developing microchips yourself.

[Vykryl]

But then you potentially get into the situation of the E.M. Forster short story "The Machine Stops", where humanity are entirely reliant on a giant machine to meet all their needs and are thus in a great deal of trouble when it breaks down and no-one knows how to fix it. (I know you said WH40K machines are self-diagnosing, but what happens if the self-repair bit breaks down?).

Just an idea to for source material for inspiration. Sometimes it helps to see how the idea was handled by others.

There was a time in the 40k background were machines almost took over through highly advanced AI. After mankind beat back its own creations it banned AI.

Think of the self diagnostics used as similar to what's in your car. There has been experiments to have your car contact your preferred shop and schedule its own maintenance. Backyard mechanics can diagnose and repair vehicles with little understanding of how the parts work through the experience of "when the vehicle acts this way its usually one if these components". In 40k most shops would resemble a backyard mechanic equipped with a replicator from Star Trek to make what parts he thinks he needs.

Main issue in 40k is when a STC file gets corrupted as they have little programming knowledge and can't debug it without a duplicate to follow as a guide. The Holy Grail for mankind in 40k is finding a complete STC library. One new file can be priceless.

There are a lot of holes in the science if the 40k universe. Still can be a good place to glean ideas from.

[wumpus]

So if a culture at, say, Bronze Age tech levels or higher encountered a steam engine, they could probably figure out how it works and how to build another, even if they wouldn't know the thermodynamic principles that underline why it works. To build and use a microchip, on the other hand, you'd probably have to be fairly close to developing microchips yourself.

James Burke made a career out of just this issue. Anybody interested in this thread needs to google him.

Hero of Alexandria (Greek, Iron age) famously built a "steam engine". It was a toy that acted more as a steam-driven windmill than a steam engine. So the question isn't "can the Greeks build a steam engine" but "how high tech of a steam engine could the Greeks copy"?

The first steam engine was the Savery engine. Whether Savery invented "the steam engine" is controversial, but he certainly either invented vaporware or at least massively developed it use. I'm sure the steel/iron quality needed to contain the pressures needed to make such a steam engine were beyond the Greeks, and they appear to beyond anyone during Savery's time as well.

The first steam engine that was economically important was the Newcomen engine. While I feel that the Greeks were *capable* of making one of these, I really have to wonder if they were remotely affordable to manufacture. They were *big* and used a ton of iron. I can't imagine a rich Greek thinking that he would spend the money to manufacture a device that takes that much iron (problem: if the Greeks can't cast iron it won't work*. And they need to cast iron *by the ton* or it won't work) vs. the number of slaves needed to turn the crank. It also needs Greeks to care about pumping water out of mines (which is about all it can do).

The Watt steam engine (which is the standard "steam engine" of the steam age) could be made somewhat smaller than the Newcomen engine, but needs even more iron (it keeps two cylinders). While it does have the advantage that it can be used for things other than pumping water out of mines (of course steam engines require so much coal and iron that you will soon need to pump out mines if you start using them) but still have all the issues of competing with pure muscle power.

To be honest, it is entirely possible that the Romans could/would have paid for this. They built massive watermills that could have just as reasonably been powered by steam (for all those places that required aqueducts and still wanted powered mills). I still think that building the furnaces needed to smelt/cast the iron to build the steam engine would be harder than the eventual design/manufacture of steam engines.

"steam engines come when it is steam engine time"
"you can make anything in the lab, building a product is the real problem"

Technology isn't driven by inventions. Technology is driven by infrastructure. You can drop whatever tech you want into ancient Greece/Rome (take the antikythera mechanism as an extreme example) and it won't change anything if the infrastructure isn't there. Good luck finding a feature in you car that wasn't in a working car in the 1920s, but you couldn't buy it in a mass produced car until the infrastructure was ready.

* maybe a blacksmith could use wrought iron to forge a steam engine. But can you imagine the cost of that labor?

[Bohandas]

It would also depend therefore on how much of the tech in the chain between the technologies of the time and the technologies needed for the device in question were also revealed.

The tech montage in Evil Dead 3: Army of Darkness for example seems pretty plausible given that Ash happened to have a carload of college textbooks with him

[Maximum77]

Yeah. I think if we found a hypergate, we might be able to activate it and use it.

[halfeye]

I said this thread title was a problem.

The actual question is can you build a copy of something without understanding how it works. The answer to that depends on the technology involved, but when you go beyond the most basic things it seems to me the answer is no.

[Lord Joeltion]

I said this thread title was a problem.

The actual question is can you build a copy of something without understanding how it works. The answer to that depends on the technology involved, but when you go beyond the most basic things it seems to me the answer is no.

Define "basic". Even "primitive" has it's problems as a standard, because it's relative both to the "average" you come from and also with the field of science involved (because not every field evolves at the same pace). Some technology is only considered "advanced" because they rely on a secondary technology that improves the original design.

For instance, transistors are pretty advanced when compared to motors (be it steam, diesel or electric) but they also work under a very simple principle: on/off. It's technically possible to build a computer with very basic tools, even more primitive and rustic than the ones needed to build an engine; what makes a computer "advanced" is the quality of the materials you are working with (crystal<metal<carbon-based<nano)

[Leewei]

I said this thread title was a problem.

The actual question is can you build a copy of something without understanding how it works. The answer to that depends on the technology involved, but when you go beyond the most basic things it seems to me the answer is no.

The knowledge of how to copy something has to exist somewhere, certainly. It need not exist within the civilization using the technology, though.

A farmer can raise a cow, create circumstances where the cow will (with a bull somehow involved) produce more cows, and milk the cow. The same farmer cannot at present create a cow from component parts.

Engineering something that can reproduce itself and produce something else, or else provide a specific task follows intuitively from the example above. Bio-engineered microbes could deposit silicon into configurations usable as circuitry. Special algae could sequester metals dissolved in water into nodules for harvesting, or to purify the water for drinking.

If a biological process such as reproduction exists, there is no reason to think that a technology would be theoretically unable to duplicate its effect at some time in the future. (Mind you, there are strong, non-technological motives for manufacturers not to do this.)

[Bohandas]

Define "basic". Even "primitive" has it's problems as a standard, because it's relative both to the "average" you come from and also with the field of science involved (because not every field evolves at the same pace). Some technology is only considered "advanced" because they rely on a secondary technology that improves the original design.

For instance, transistors are pretty advanced when compared to motors (be it steam, diesel or electric) but they also work under a very simple principle: on/off. It's technically possible to build a computer with very basic tools, even more primitive and rustic than the ones needed to build an engine

Yeah, you could potentially make a babbage engine out of unalloyed copper

[wumpus]

Define "basic". Even "primitive" has it's problems as a standard, because it's relative both to the "average" you come from and also with the field of science involved (because not every field evolves at the same pace). Some technology is only considered "advanced" because they rely on a secondary technology that improves the original design.

For instance, transistors are pretty advanced when compared to motors (be it steam, diesel or electric) but they also work under a very simple principle: on/off. It's technically possible to build a computer with very basic tools, even more primitive and rustic than the ones needed to build an engine; what makes a computer "advanced" is the quality of the materials you are working with (crystal<metal<carbon-based<nano)

I mentioned transistors earlier in the thread. I've heard that there is an earlier patent on transistors from the 1920s, and the "inventor" actually had at least one working transistor. Unfortunately, he was never able to get them into production because he didn't sufficiently understand the underlying technology (you need very specific impurities in various parts of the transistor) he couldn't put it into production (the "factory" kept producing too-pure silicon).

To determine how a transistor works, you need a strong understanding of the crystaline structure of silicon as well as the effects of "doping" (intentional impurities). It is possible that had chemists analyzed enough working transistors, they might find the "magic difference" and make transistors while skipping that step, but that's not what happened in history.

The point here is that you need to understand things 'enough'. What is enough entirely depends on the specific technology (and really doesn't depend on how "high tech" something is, just how close the "good enough" understanding is and how close it is to the underlying reality (which presumably requires god-level science, and we are just picking pebbles at the seaside). Newcomen made an economically viable steam engine, but it had obvious flaws and ground to a halt after time. James Watt was tasked with repairing a [Newcomen] steam engine model (at a college), and once he understood what was happening he found the major flaw, corrected it, and the result was such an improvement he is so associated with the steam engine that most people think he invented it in the first place. You can make a steam engine with a rough understanding of how it works, but you can make a really good steam engine with a *good* understanding of how it works.

The Wright brothers invented the wind tunnel but didn't understand how badly it scaled. Thus the famous "Wright Flyer" (first airplane) was designed assuming that air had the viscosity of molasses (I'm assuming this is a wild exaggeration, but it is in all the books). This was still "good enough" and the airplane flew. Eventually they figured out the problem and had to build larger and larger wind tunnels.

Until science is finished and the universe is understood perfectly, it is pretty much impossible to "perfectly" understand a technology. There is only "good enough", "barely good enough" (like the Newcomen engine and Wright Flyer"), and "not good enough" (like the non-reproducible transistor). I'm not suggesting that I believe science can ever finish (and don't assume you can solve the three body problem even if physics deduces the "theory of everything").