So I was thinking about minimum viable population levels for a colony and it hit me that there are two industries that probably can't be made to work at small population, but that you need to have a high tech society. (I'm not asking about minimum genetic population, but minimum needed to be able to make all the stuff you need for a high tech civilization out of local resources. Two of the more involved parts of the supply chain are mentioned here, but there are other industries with similar issues. I'm interested in the local supply chain needed to make a viable colony.)

The first industry is semi-conductors. You need them for computers and embeded smart devices. But it isn't a cottage industry. A modern "fab" costs two billion and the cleanliness rules are staggering. Plus, even if you do have a fab, switching between products is difficult and prone to start up problems. (Yield of working chips can be very low at first.) And you need a lot of different types of chips and you can't repurpose a production line used to make memory chips into one making cell phone chips. (The process is really different and the machines aren't the same.) So this is going to have to have multiple production lines, multiple groups of production workers and different groups of engineers, since an RF engineer is probably not that good at designing microprocessors. You could use a library of proven designs, but that means that your tech stagnates and frankly building some of the tools (vacuum systems and ion implanters) is going to be a very tricky industry in its own right. Figure you probably need thousands of people working at this full time to make it even marginally possible.

The second is the pharmaceutical industry. It has the same type of problem, there are thousands of products and the process to manufacture them can be very cranky and they need different processing equipment. So again you need multiple production lines, different groups of production workers and different groups of researchers. And bio-med research is even more specialized than chip design. Again, they could have a library of processes to make known drugs, but then again progress stagnates. This industry is even larger than the semi-conductor industry so I think it would take low tens of thousands to be able to make all the different drugs required.

So from this, I think we end up just two industries requiring thousands of people just to keep the wheels turning which points to a population in high tens to low hundreds of millions.

So I guess Mars, the Belt and the Moon are going to be dependent on Earth imports for a long time.

  • 2
    $\begingroup$ The Aegis system allows a Ticonderoga-class cruiser to be operated by a single person. The normal crew complement of a Ticonderoga-class cruiser is about 430. Just sayin' $\endgroup$
    – cobaltduck
    Feb 10, 2017 at 21:54
  • 4
    $\begingroup$ Possible duplicate of What is the minimum human population necessary for a sustainable colony? $\endgroup$
    – Mołot
    Feb 10, 2017 at 21:55
  • 2
    $\begingroup$ Can you talk a bit to what "high tech" means? I find that many people's definition of "high tech" is synonymous with products which require an incredible amount of highly specialized workforce. I'd argue every single STEM job out there demands a huge population, just by the nature of the jobs we apply STEM to these days. Also, as another example, steel production vastly benefits from economy of scale. As does aluminum manufacturing. As does modern farming methodology, for that matter. $\endgroup$
    – Cort Ammon
    Feb 10, 2017 at 22:07
  • 2
    $\begingroup$ Welcome to Wordbuilding! Can I ask: what's the actual question that you're asking? If it's 'what is the minimum population?', then I'm afraid that 1: it might be a duplicate, as the other question has answers that also answer this and 2: you seem to be almost answering your own question ("thousands of people") within your question, which is bordering on no actual problem to be solved. However the actual text you've got seems to be leading to a few other potential questions, so can you be more explicit about what you're asking? Thanks $\endgroup$ Feb 10, 2017 at 22:26
  • 4
    $\begingroup$ If "High Tech" can mean "Futuristic High Tech", then the required population size can very well be 0 - robotics can take care of that :) $\endgroup$
    – Alexander
    Feb 11, 2017 at 1:44

5 Answers 5


Overall you have right thought about the situation, but incorrect conclusion. So, basically, it is a right question and a very very good question which has to be asked more often.

The "amount of people needed to support all the technology" is a part of a bigger question, and more general question - how big have to be a colony to be self-sufficient.

The answer is different for different tech levels. It is different not only in terms of low tech level needs one person to make stone knives and axes, the number of people varies also and for higher technological levels, if we compare your current productions with 1950x, 1900x.

The need to live in space sets a minimum for a technological level which has to be with rolling wheels.

Elon Musk (SpaceX) projection goal for Mars colony is about 1 million people, I would say that 20 million would be better, wheels will roll easier.

So far similar to what you describe, but as long as we include the Belt and thus the space with microgravity, everything changes a bit.

One of the benefits of microgravity and space is the energy costs, it costs less effort and energy to generate it, it is much cheaper than on earth or on a planet like Mars.

The reason for that is simple, no gravity means that a 1 um aluminium foil with area 1 km2 will keep its form without any support. You make a dish from it - it stays a dish, you make an elephant it stays in elephant form. It does not rust, it does not oxidize. The only thing it gets some holes from micrometeorites. (not totally true, but it just oversimplified picture of the difference)

The situation is not achievable on any planet or moon, but in space in microgravity is perfectly fine.

The second factor of the cheap energy, it does not ends (solar energy), even 3 a.u. it can be considered as cheap and endless. It does not depend on day/night, clouds, winds, you do not need accumulators, you do not need to dig for it - it is constant and almost the same power 24/7/365.

And now what?

Let's see Intel Fab42 enter image description here

Area about 200'000 m2, and I guess 50m height is enough - so potentially the Fab can be fitted into volume of 0.01 km3

So one cubic km can contain 100's of such fab's and different fabrics/manufacturers for the technological loop to be self-sufficient.

A 1km3 construction in microgravity is not a lot, it is not a problem compared to Burj Khalifa 829m height in 1g gravity. The construction can be light weight as it basically has no stress from gravity, and it can be stiff only in places where it is needed(fabric in a container).

How many people, intro

Development - yes it needs people, but if you do not have people for development it is not a problem, you will just have the same technology each year, and they will say it worked for my grand- grand- grandfather, so it will work for me. There is no dependence on development, they do not have to get more from the same surface, because they can just grow and scale the production according to the needs.
But if you need development, then yes, it needs a lot of people(but there are some solutions to the problem), but 20 million at current technological level is enough when everything is organized a bit more effectively than it is now.

The efficiency of production - also not a factor if they do not try to compete with someone else, and make the things just for their own consumption(and survival). If there is only 10% of good things(chips whatever) which pass quality control - it is not a problem when energy is order's of magnitudes cheaper, the scrap goes in the recycler.

Energy - as energy is not a problem then it is not a problem to keep a big factory(same Fab42) up and running 24/7/365 producing million of chip day, and do that just for one human, for him to have ability just to replace his phone, once a year in case the phone is lost or broken.
Used or not the energy never ends, and you can't limit it, you can stop convert it into useful work but that's all. If you do not convert it to useful work it flies into space to aliens.

How many

The main factor will be, yes, how many people are needed to keep the thing operational and how many are needed to repair the thing. (actually, it is the same stuff but...)

At the moment high tech factories have pretty good automation, the human labour is in repairing the machines and the robots and in checking and quality control. When you do not care about humans who have to work somewhere (because you do not have those resources, those labourers are the initial problem - no people, no problem), about efficiency, about high percentage of passing quality control production, about energy for recycling etc - then in that case it makes sense to put a robot arm in a place where we have humans at the moment, even it if kinda costs more and is slower and not so accurate - you just put it because you do not have a human to operate, but you have an ocean of cheap energy to compensate for possible flaws or the cost of the solution and when the only desired property is to not need humans to operate the facility.

In those circumstances, I guarantee you most of the workplaces can be replaced at today's technological level. Repairing - do not repair, just put a new machine, and send the old machine in recycling(if there is no technology for repair, which is relatively easy can be done).

So 10 belt miners will have 10 cubic km mining complex which includes all factories needed and a huge foil mirror to supply the thing with solar energy - and they will be self-sufficient 100%. They can't create the technological seed, but they will be perfectly capable of managing it.

I did not mention a lot of other factors, such as example - lab equipment which can produce the same 7nm chips or other things - usually it is not that big, it is kinda sophisticated at its level but not so much, it is not very efficient especially for mass production - but as long as we do not care about efficiency it is not a problem, automation of such labs is possible etc

So answer is No, they will not depend a long time from anything, as a first technological seed will be produced. And the technological seed will be less than 1 km3 and it will need less than 100 people to manage it.


The number balloons into the billion quite quickly. It is impossible to give an exact number but we can try a fe estimations.

Just to prevent stagnation you are going to need a constant turn over on specialists, you can't have just one heart surgeon you need hundreds just to insure innovation and skill. Then you have to ask how many people do I need to to provide enough work for those hundreds of heart surgeons now consider how many bus drivers and general practitioners that population needs. then you have to do that for a million other highly specialized jobs,now of course you will get some overlap in support but then you consider the material needed to support them , how many hundreds of thousands of people are needed to build one MRI machine, to mine the rare metals, design the software, manufacture the plastics and electronics, then you need technicians to run it a infrastructure to support it.

As for your two industries image how many people you need to mine the rare materials used in the chips then how many farmers and doctors and teachers you need to support all the combined people, according to the worldbank the global average is a little over a thousand researchers per million people, for more research minded countries you can get as high as 5-6000 per million but those countries are not self sufficient so I would not go much higher than the average. Now consider how many researchers you need to keep all of technology moving forward.

Current estimates are for about a 6-7 million researchers worldwide, even if you halve that number you still need 3-4 billion people to support them.


Depends on how high tech - extreme automation and hypothetical ultra-high-tech manufacturing technologies (incorporating nanotech fabrication etc.) could make the number arbitrarily low.

Pharmaceuticals could be greatly simplified by really advanced biotech. If you really understood the body, at every level from molecular to whole organism, and had the computer capacity to model it, you wouldn't need all the complex testing we do now -- the computer models would just tell you "this molecule will have these effects".

As for the actual manufacturing, well, you could just splice the precise DNA sequence to make that molecule into some organism, and it would produce it for you.


With a couple of changes to our perspective, it could be quite a bit lower than we would assume.

We wouldn't need anywhere near as many different semiconductors as we have to be high tech. A lot of specialized chips could be replaced with general-purpose computing chips, perhaps with some inefficiency in size, speed, power usage, etc., but traded off against efficiencies in manufacturing and logistics and the number of people needed. If we didn't design everything to be disposable or for planned obsolescence, we wouldn't need to keep making new chips with trivial differences that in the end do the exact same things for the end-user. This needn't mean stagnation, innovation could be more significant though less frequent.

Pharmaceuticals are largely driven by profit and guesswork. We could probably maintain tech levels but with a lot less complexity and people needed if we had better tech to know which medicines were needed (instead of the "try this and if it doesn't work we'll try something else" approach). Also if we were willing to accept some tradeoffs - many drugs are effective at treating multiple problems.

In addition, a lot of the population does things that are not strictly necessary for maintaining a high-tech society but are a convenience or a luxury. While not everyone is going to want to be a research engineer, if we can reallocate over time and increase the ratio of population working on research, production, and development or the support thereof, we would need a lower population to sustain it. On a space colony, this would be a must.

The way our economy is structured, with extreme specialization, it would most likely take millions or tens of millions or more. But if we made some tradeoffs for efficiency and generalization and restructured the economy to support it, then a population in the tens of thousands or hundreds of thousands (or possibly lower) could potentially work. (Once infrastructure is setup and automation applied where possible).


The USA currently is just over 300 million and is essentially self sufficient and technologically advancing. Yet even in the US there is tons of useless activity from fuzzy dice to monster trucks and celebrities. A lot depends on how controlled, focused, directed the population is. Sure they engage in trade for raw goods they don't have but if they were the only people on the planet they would just spread out and collect the resources themselves. My guess is you probably only need about 2 million or so with a tech base. Consider Asivmos robot novels with a few people supported by a lot of tech. A decandant society but is there some rush for development? Change doesn't have to be breakneck pace.

  • 1
    $\begingroup$ I am not sure the USA is self sufficient. I am sure most of the highly skilled workers they use nowadays are immigrant. $\endgroup$
    – L.Dutch
    Aug 6, 2019 at 7:27

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .