Tuesday 21 March 2017

Interstellar Trade Is Possible

In this post, we will detail a method for developing interstellar trade using near-future technologies and commercially realistic requirements. We will then look at the various outcomes, challenges and development models that will follow the first interstellar operation.
There is now a Summary at the end of the post.

A tough task
Travel between stars is hard. The distances are measured in trillions of kilometers and the space between destinations is not really empty. Attempting the crossing at interplanetary speeds is ludicrously slow; the only way is to reach velocities measured in percentages of the speed of light. Even then, travel times might measure decades to centuries.
Our nearest interstellar neighbour is Proxima Centauri, at 4.2 light years (40 trillion kilometers). Then it is Barnard's Star at 5.9 light years and Luhman 16 at 6.6 light years. There are thirty-six stars within 12 light years, in twenty-three systems, of which only seventeen are not brown dwarfs.
These will be targets of our discussion.

The distances are, again, extreme. They are a significant hindrance to travel, but even more so for trade. This is because we expect returns on investment to match the time we are separated from our money. If I put $10000 and want to earn $1000, I can put it into a fund that pays me back in one year. If I want my money back sooner, I'll accept a lower return. If the fund asks to keep the money for 10 years, they better double my money.
How to calculate how much a 'promised' amount is worth today. Discount rate is how much the investor want to earn per year.
These expectations are problematic for interstellar trade, as there is an lower limit on how quickly your money can be returned to you. The initial investment is great: pushing an all-in-one spaceship (can depart and return from a star, set up its own profit-making base, maintain it long enough to make a profit without any external help) to most of the speed of light is done at enormous cost. The biggest cost is propulsion. Lowering the propulsion cost means longer travel times. Some designs go all the way and turn the spaceship into an Ark where humans grown and die with the task to increase the chance of success of the mission... these designs are frequently measured in thousands of tons of the most advantaged technologies. This also disregards the problem of finding competent engineers willing to die halfway to the destination in a tin can.
Sorry, human travel to other stars will have to wait.
Just what special product extracted or produced around another star can justify such extreme waiting times and investments? Even if the operation makes a profit overall, leaving my money in a fund at home could turn my money over several times and become a more interesting option. Why should I bother?

At least, that is the common understanding of the matter. Here is another perspective. 

Problem bounding

How long would you be willing to give up your money for an interstellar operation?
Merchants willingly spent years on the Silk Road.
The longest modern delay between paying for something and getting it is 100 years: certain bonds issued by governments or large corporations have a very long maturity date. 10 and 50 years bonds are much more common. Even so, there are options to get back some of the money paid for the bonds in shorter-term increments. 

More substantial operations, such as the construction of certain canals, walls or historical monuments, have taken even longer to complete. Some projects spanned centuries. However, they offered intermediary benefits or had non-commercial goals. The Great Wall of China took thousands of years to complete, but the completed sections served their purpose in the meantime. The Great Pyramid of Giza took 20 years, but it was mostly a funerary decoration. They are not a good comparison to interstellar travel, as there is nothing to 'reach' between destinations, and no benefit from being 'halfway there'. Its an all-or-nothing endeavour with few current comparisons.
1.12km wide, 175m tall.
A more appropriate example is the Three Gorges Dam in China. It took 17 years to complete, with no intermediary benefit. Although, it was a state-run operation and mostly funded by the country's Central Bank, not private investors. 

So, if we want to trade across interstellar distances, we want to do so on timescales investors will agree to. This means that the return on investment must be made before 50 years. We will also look at 20 and 10 year options. 

Within this timescale, the spaceship must reach the target, conduct its activities and deliver goods.

Travel times

If we want to reach Proxima Centauri and return within 50 years, we'd need a spaceship that travels at a minimum of 16.8% of the speed of light. To do so in 10 years, you need an average velocity of 0.84C. To reach the furthest star on our list, Gliese 1016, withing 50 years, a minimum velocity of 0.48 C is required. 
Attempting to arrive and return on shorter timescales requires proportionally larger velocities. Some destinations have minimum travel times imposed by the speed of light. You cannot reach Barnard's Star and return in less than 11.8 years, for example. 

Travelling closer to the speed of light to shorter travel times is necessary to trade at more human timescales. However, approaching the speed of light has a hidden cost: relativity. 
At 10% the speed of light, the relativistic effects are negligible and the kinetic energy equation (1/2*m*v^2) accurately describes the propulsive power required. However, the closer you get to the speed of light, the apparent mass of the spaceship will increase. This is described by the Lorentz Factor. 

At 10% of the speed of light, the Lorentz factor is a negligible 1.005: the spaceship would only require 0.0005% more energy than expected to reach that velocity. At 20%, it s still only 0.002%. At 80%, it is 1.667. At 90% it is 2.29. At 99%, a Lorentz Factor of 7.089 means that the spaceship requires more than seven times more energy than indicated by the kinetic energy equation.

It is therefore evident that there are quickly diminishing returns for trying to travel faster. A spaceship that travels at 90% of the speed of light instead of 80% only reaches its destination 10% faster, at the cost of twice the kinetic energy.

We also have to consider the Time Value of Money. Lowering the travel velocity lowers the initial cost but extends the duration of the trip. The longer the trip takes, the less the promised amounts of revenue are worth.

If you want to earn $1000 out of your $10000 investment next year, then your Required Rate of Return is 10%. If the investment takes 2 years, then you'd want 10000*1.1^2: $12100 in money back. Three years, and you'd want an extra $3310 instead of just $1000. In other words, the longer it takes to complete the trip, the more money that has to be promise to investors. 

One way to reduce travel times is to fit bigger engines or use more powerful propulsion systems. This costs more money, right from the start. If you want to travel 20% faster, you might need a 20% bigger initial investment. An investor would have to put in $12000 instead of $10000. He'd want $3200 more after the first year, instead of just $1000. 

A balance has to be struck between the motivation to reduce travel time and the larger up-front costs.

  • Time cost of travel: RR ^ ((2 x Distance + CDT )/ Velocity)* Initial Investment
  • Propulsion cost of travel: 0.5 * CJ * Mass * Lorentz Factor * Velocity ^2
RR is the required rate of return. CDT is the Colony Development Time, which is the delay between the seed arriving and the first payloads returning home. CJ is the Cost per Joule delivered the the spaceship. It is calculated differently for external propulsion systems or propellant-consuming rockets. Lorentz Factor is as was explained above. 

Here is a simple example:

Imagine the fixed cost for the operation is 10 million units of currency. Distance is 5 light years. Spaceship is a laser sail of mass 1kg. Cost per joule 0.5 million units per terajoule (roughly running a 10GW laser for 1 month). RR is 10%. CDT is 1 year.

Using those equations, we find that the optimal velocity is 0.22C. For a ten times high initial investment, the optimal velocity is increased to 0.32C. For a ten times higher cost per joule, the optimal velocity is 0.16C. 

The solution

If we want to go fast, we would need investors to accept huge up-front costs. If we want to go slow, investors will give up in favour of more immediate sources of income. On top of that, long travel times increase the chance of failure, such as radiation degrading electronics or collisions poking holes in the spaceship. Compensating for failures with a human crew rapidly makes the spaceship an Ark massing thousands of tons, possibly as expensive to launch as a fast probe.

The StarWisp concept.
Instead, we should go small. 

A tiny micro-spaceship is cheaper to accelerate to large fractions of C. Because it is small, we can afford to build many of them. Sheer numbers and short travel times compensate for the failure rate. There is also a hidden advantage: micro-spaceships can piggy-back on existing infrastructure or use off-the-shelf technologies. After all, investors are unlikely to wait through years of R&D and testing cycles before anything happens.

A small spaceship will reach the destination star-system, enter orbit and latch onto an asteroid or comet. It will use the resources available to build more copies of itself. Once a critical number of components is reached, it will start specializing the members of the 'colony' into roles such as energy collection, resource collection, communications, propulsion and so on. This colony continues to grow  at exponential rates. It quickly reaches the size of the 'all-in-one' spaceship designs mentioned above, with similar functionality.

This ability to self-replicate and then specialize takes cues from the natural world. It can beat the time/velocity constraints set by traditional modes of interstellar travel.

How small are we talking? 

A few grams at most! Three main versions could exist.

Binary fission of bacteria
The first is based on bio-technology. Literally taking from nature, we could engineer hardened bacteria with data encoded into their DNA. These bacteria are frozen and vacuum packed into a nutrient jelly and incubator. Upon arrival, the hardened bacteria are unfrozen to produce the bio-tech cell-based replicators. These can afford to be much more vulnerable to damage if it allows them to consume local resources (water, rock) using sunlight for energy. Specialization requires a 'trigger' that switches the replicators from building more of themselves to building artifical machines out of cellular products such as muscle fibres, cytoskeletons and plastics. Some might be able to etch microprocessors out of silicon and create electronic machines.
Bacteria forming 2D crystalline structures
Bacteria forming honeycombs
At a rate of one division per hour, with external mechanisms clearing out waste products and supplying fresh resources, a 1 gram 'colony' can reach 1 ton in 19 hours, 1000 tons in 29 hours and 1 million tons in 39 hours. Bacteria consume about 1kW/kg when growing rapidly. The energy requirements of such a colony grow from 1W to 1TW. This would require that the colony slow down its growth to dedicate specialized to cells to producing a solar panel. Some techniques, such as detaching part of the colony to orbit the star closer and beam the energy back, or ejecting part of the comet/asteroid as propellant for propulsing the colony to a more favourable orbit, can help reduce the non-replication fraction of the colony.

Nonetheless, the expansion of a gram-sized colony of replicators under favourable conditions taken an insignificant fraction of the interstellar travel time.  

The second is based on micro/nano-technology. Very small machines contain all the data electronically. They are currently outside the reach of modern technology, but it is not a far stretch. They would be more vulnerable to damage than self-repairing bio-technologies, but are more efficient and develop faster on the same energy and resources than cells.

The third is theoretical atomic machines. Smaller than even nanomachines, these can manipulate matter at the atomic level. The advantage is that the 'spaceship' can be nothing more than a handful of molecules massing less than a milligram. The disadvantage is that it would be very difficult to store data on what the machines should do in a few molecules. 

For the rest of this post, the cells or machines will be referred to as the 'seed'. This seed 'grows' and produces an 'ecosystem'. This ecosystem eventually returns products to the home system.

Examples of such an approach to interstellar travel do exist. Robert L. Forwards' StarWisp. It involves a 1 kilogram laser sail, propelled by microwaves to 10%c. Another is Geoffrey Landis's one-ton probe that uses a laser dish to power an ion engine to high exhaust velocities. 

To return products to the home system, the ecosystem produces a replica of the propulsion method that propelled it in the first place, but at a larger scale and dedicated only to interstellar travel. 

How it works

We will now look at the entire operation as a series of steps to take.

Step zero is to assess the current situation. As time goes on, humanity's economic, technologic and politico-legal status become more or less favorable for interstellar operations. They are best attempted when it is easy, affordable and profitable to do so.

The first step would be to find money for setting up the first operation. This is different from interstellar exploration, which would be funded mainly by governments and research bodies, and hampered by the fact that it is much easier to just build a bigger telescope than to send a probe to another star. Funding interstellar operations involves attracting investors and laying out a plan that produces tangible results at minimal cost and short timescales.

Next, the interstellar operation delivers the 'seed' to its destination. Interstellar travel is explored more in depth in the next part of this topic. This is the most 'hands-off' part of the interstellar operation. It ends when the self-replicating colony has grown to the point where it can divert resources to sending a signal back home.

The third step requires tough decisions to be made regarding the colony's future. With a several-year time lag, the investors can order the colony to concentrate on growing, on industrial production, on spreading across the system, on building a propulsion system to return products to the home system or sending copies of itself to other star systems. The tactics, strategies and competition the colony faces will be detailed in the third part of this topic.

The fourth step is critical. It involves receiving actual tangible products from another star system and turning a profit. How to do this and how it will affect the wider economy are critical.

In the case of a 10 year mission, we want returns higher than even the riskiest governments bonds, and equal to the corporate investment rates. This means a 10% return per year or better. If investors hand the company $10000, they want  $25900 or more. This is not an extreme requirement, considering that the investors only send out a gram-sized seed and get back tons of rare elements on their doorstep at no extra cost. 

Our situation

Could we even fit a spaceship on something the size of a microchip?
Today, we are unable to mount an interstellar operation. Although we have the bare minimum of propulsion technology to send gram-sized payloads to another star, the cost of doing so is several times the world's GDP. There's also the matter of creating a workable, reliable self-replicating technology. The biggest hindrance to the project, however, is that the money required to mount an interstellar operation gives much better returns staying at home. 

Nothing, absolutely nothing, that comes from another star system is cheaper than what can be extracted or made on Earth.

Schemes to build massive solar power stations to fund interstellar travel by selling electricity misinterpret the nature of Supply and Demand
That statement will likely remain true for a very long time. Centuries, likely. 

The development of more advanced technologies does not necessarily mean that a profit can be made on interstellar operations. Self-replicating seeds can be 'planted' right here on Earth to provide inexpensive exploitation of even the most recalcitrant mineral vein or most rarefied elements from air and oceans. Energy can be provided from uranium extracted from ocean-water uranium or orbital solar satellites using the same construction techniques required to build self-replicating colonies on other star systems.

Cheaper, more powerful propulsion technologies will open up the resources of the solar system before making interstellar travel accessible. At current rates, it will be tens of thousands of years before humanity lacks the raw resources to continue expanding. Whether it is the billions of tons of uranium, the trillions of tons of heavy and rare metals in asteroids and the practically infinite amounts of iron, silicon and carbon on the planets, our needs are covered!


This is a false view of the realities of humanity's growth. 

The energy and resource consumption rates are not being driven by population growth anymore; more precisely, they exceed what can be explained by larger populations. Developed countries are competing by producing for ever more energy-demanding markets, such as smartphone factories and computing. Developing countries are doing the same in voracious industries such as steel, bulk chemicals or food production. Today's 'under-developed' countries are starting their way up a mountain with an ever growing peak titled 'kWh per capita'. 

The fastest growing economies are based on the most energy-intensive industries 
Our need for energy increases every year, and the rate of increase is also growing. We currently need about 15TW of power to fuel the planet. This has increased five times since the end of WWII, twenty-five times since the start of the 20th century. It will likely increase by as much or more by the next century, to 100TW. 

Coal and other fossil fuels have historically absorbed most of the increases in energy demand. Renewables (solar and wind) are the fastest growing energy sector. A significant portion of today's energy demand is produced by nuclear power stations. Efforts by the China Atomic Energy Authority are a major contributor to this development.

Will we have enough?

Fossil fuel reserves in Zettajoules. 1ZJ powers a 100TW civilization for 31 years. 
That is not a good question. We will always have enough energy. 'Peak oil' is a long way off, if we consider the massive reserves held in oil shales and other unrecoverable sources. If it is not oil, nuclear energy will last for thousands of years. If we convert reactors to use thorium, there is enough for tens of thousands of years. Then, there is the Sun. It will always be there!

Also, energy alone is a rather useless measure for our purposed. What matters is costs. 

The cost of producing energy is related to the cost of making power stations. The billions of dollars required for a nuclear reactor or the thousands per solar kWh are an addition of material, financial, human and legal (insurance/safety) costs. All are variable except for material costs. We can be certain that these can only increase over time relative to the others. Whatever the production method, they cannot compensate for the fact that there will be less raw resources over time, held in deeper and harder-to-reach locations.

The most important of these raw resources are rarer elements, such as chromium for stainless steel turbines, cadmium in nuclear control rods, indium for solar panels. Many, such as copper, are majoritarily met by recycling. 
Recycling allows us to forever have some production capacity, but it cannot compensate for increases in demand. 

The next step is moving into space.

Solar power stations, rare elements mined from asteroids, lunar industries. We won't use them because we have completely run out of power or rare elements on Earth. We won't even wait for them to be significantly cheaper options than building another nuclear power station or platinum mine. No, they are viable as soon as they can make a better profit than their equivalents on Earth. Making a profit can be done at higher overall costs, through cheaper cost per ton or kWh output.

An example is oil. It cost on average $60 to $80 per barrel over the past ten years, despite the recently artificially lowered prices. It cost less than $10 per barrel 30 years ago, less than $3 dollar per barrel 50 years ago. Despite producing much more than before, the increased energy consumption and the fears of oil running out has made prices skyrocket. 

This will happen to all energy sectors: wind, solar, nuclear... Even if the cost of increasing the surface area of a solar panel installation becomes dirt cheap, or if nuclear fuels become as widely available as gasoline, the price of electricity will still have to reflect the increasing value of rare metals and elements they are made of, and the cost of opportunity cost in covering arable land and habitable spaces with power stations.   

But what about going interstellar?

In macroeconomics, there is the concept of the 'long run'. Simply, it is looking at industries and companies over time periods long enough that even fixed costs become variables. 

Long Run Average Cost curves allow us to explore the concept of diseconomies of scale too. 

Using Long Run Average Cost, we can look at resource consumption, energy production and economy growth at all scales and over long time periods. For example, it allows us to disregard the effects renewable energy replacing fossil fuels, or a move away from rare and hard to find metals towards common but hard to make graphene in electronics. 

We know that the cost per kilo in orbit will fall. The cost per solar satellite kWh will become lower than the same kWh on the ground, because it will not have the same restrictions. The profit margin on space-based industries and interplanetary operations will increase. In the long run, the entire Solar System is one big competing economy where growth equals cheaper prices, due to economies of scale. It would not cost much more to invest in an asteroid mining operation than to build a lunar factory or send robots to Uranus, relative to the output. 

But, even on the level of a Solar System, humanity will meet diseconomies of scale. 

The number of asteroids with rare elements is limited. Mining planetary surfaces will reach profitability ceilings quickly, much faster than on Earth due to higher base costs. Advances in propulsion technology will help to match prices across the Solar System but only slow down the relative scaling of resource consumption. 

So when will interstellar operations become viable? Is it when they start being profitable compared to opening up another mining operation on Venus or extracting tons of copper from an asteroid for the next solar power station? 

Even earlier!

Property claims travel faster than rockets. The domains of states and nations travel faster than the speed of light! If, say in 100 years, the United States establishes an inhabited colony on Callisto, will it not claim the entire moon for itself? If China invests several billions in automated mining machines around Neptune and its moon Triton, and protects them with warships, will it not consider it part of its sphere of influence?

Building artificial islands to secure your claim on the area sounds a lot like ringing space stations around the moon you're interested it. 
The Outer Space Treaty does not allow anyone to 'own' planetary bodies. It doesn't say anything about owning the space around them. Even if that loophole is ratified, you can prevent anyone else from entering your orbit for 'safety reasons'. You can ask anyone to stay outside of a certain radius of your position due to a 'risk of collision' or simply apply a version of the exclusive economic zone to space settlements. 

Depending on how the law is re-worded or clarified in the future, the entirety of the Solar System, out to the Oort cloud and the Kuiper belt will be divided between corporations, nations and states and 'owned', either as property, exclusive economic domain or through less legal bullying or loophole regulation enforcing. For example, if I shoot out a micro-sized spaceship, like the seed described above, at an icy Trans-Neptunian object, and turn it into a flying rocket fuel depot, I would have enough rights to the object and its use to prevent anyone else from settling a second colony on it. 

This level of appropriation means that most of the Solar System can be 'tagged' by tiny spaceships, quickly, with the most powerful actors taking the lion's share. 

How does this relate to interstellar operations?

Well, if the entire Solar System is owned by somebody or another, and the growth limits imposed by finite resources encourage state and private actors to bully new entrants out of the game, the only way is out. 

A group of investors trying to beat the market in a fully owned Solar System will have a hard time trying to do so while paying rent and fees to existing market players. A company trying to expand vertically cannot fully control its suppliers. If a monopoly is put into place on a specific resource, it cannot be broken from the inside.

Nonetheless, these conditions might take centuries to appear. This is a difficult truth for science fiction writers, as all of technology, science, culture, social norms, politics and even languages change considerably. Trying to recreate their movements and still relating it to a modern audience is a very, very difficult task. Most of them want to write a story with a science fictional setting, not a setting where things happen. As reader, you want to learn the characters, not try to decipher the setting through the eyes of someone who finds it normal. 

Making money

The interstellar operation's biggest variable, upon which depends profitability, cost and everything else, is time: t
ime spent travelling to another star, time spent developing an industrial ecosystem, the time it takes to receive interstellar products.

Taking longer to complete objectives has to be compensated by higher or more reliable output. The output of an ecosystem set up in another star system is very likely to be proportional to the size of the colony and the fraction of its mass dedicated to non-replication roles. This means, that unless specific instructions are given to prioritize industrial production or colony expansion, the output will start small and increase quickly.

With the small initial output, it is best to prioritize what products are returned home. It has to be easy to make, so no complex microprocessors. It has to have a high value per kilo, as the output will depend on mass and the energy cost is fixed.

The first returns from an interstellar operation will be bulk quantities of the rarest elements. It will be whatever the Solar System was running out of when the seed was launched, whatever high-priced item that was expected to continue increasing in price. 

The ISV Venture Star is an example of an interstellar vehicle using a combination of lasers and nuclear engines.
Due to the relatively smaller quantities involved, the products will likely depart and stop at our Solar System using the same nuclear-powered engine. This minimizes the cost of capturing the products into a planetary orbit for sale. 

Later, larger quantities of material would need much bigger engines. It might not be feasible for a self-replicating colony that started out of a 1 gram mass of cells to produce reliable uranium centrifuges en masse a few months later. It would lack the ability to fuel up goods-laden spaceships for the a return journey. Accelerating them outwards can done with local solar-powered lasers and it becomes cheaper to build a laser around the Sun and brake the spaceships in the same way on our end. Discussion of interstellar travel methods in part 2 will cover the topic of returning products home.

As the products start returning home, and the interstellar operation starts making outrageous profits by breaking monopolies or crashing price bubbles on scarce resources, a certain dynamic process has to be initiated to continue making profits.

This process is the slow communications between the investors and the seed colony. Communications are done at light speed, so it takes between 4.2 to 12 years for a message to arrive at the colony. The same delay is required for a 'message read' confirmation to return. 

What do the investors do with these communications?

Everyone hears the colony's signals... but it might be encrypted.
Well, they dynamically compare the colony's output and growth reports to how the market is evolving back home. They update it on what to stop collecting and what to focus on. At a certain point, it would make sense to start developing the colony's complexity than its size. If it can process ores into finished chemical products, or even shapes and machines, then it can increase the value of its output without increasing the size of the payloads returned to Earth or the energy consumption of its launch apparatus. 

New information can be sent, allowing the colony to follow technological trends and continue producing relevant components. When the highest value asteroid resource of the colony's star system are nearing exhaustion, investors might point the colony towards the planetary bodies. It would have grown large enough to compensate for the energy cost of moving supplies up and down the gravity well. The reward would be elements that are rare even on asteroids, such as uranium and other fissile fuels. Some industrial processes require gravity, other are extremely polluting to the environment or plain unacceptable on inhabited surfaces. These can performed without care.

The difficulty would lie in anticipating the market trends and compensating for its movements in expectation of an interstellar delivery. For example, if indium is in demand, its price might drop in response to an interstellar colony announcing a delivery of several thousand tons of it. The investors can try to trick the market by labelling the delivery as something else. Or, they could have looked ahead and understood that indium flooding the market would make solar panels cheaper to make, which actually increases demand for selenium. So they sent instructions to the colony 4 years ago to closely follow up the delivery of indium with a big ball of pure selenium 20m across. 

Thankfully, trends over 4-12 years are considered 'long term'. They move much more slowly than spot prices. Their volatility is further reduced by the fact that interstellar operations deliver products at a rough price of zero above amortisation and product recovery cost. 

Since the useful life of an interstellar operation is nearly infinite, and it can boosted every few years with new seeds like a shot of corrective vaccine, the amortisation cost is purely arbitrary. The recovery costs are drawn from the profits made from the deliveries. Since the majority of these costs are spaceships that go and dock with the payload or an array of lasers that brake it into the solar system, they can be considered investments. The operation is investing in an increased capacity for handling larger payloads. 

An investment model

Do the investors sit doing nothing in between payload deliveries? Are physical products the only thing they have to sell?

Not at all!

In a modern economy, the investors will be able to make money before the first signal arrives from the colony and have profits that far exceed their sales revenue.

Bonds, futures and intangible assets are vital to an interstellar operation.

First of all, consider the steps involved in mounting an interstellar operation. While many will hear of the attempt, only a fraction of people will have money available for investing in the project. A small percentage of those people will actually put money on the table.

Will that percentage not increase once the probe is on its way? Will it not increase with every year the seed spends in space not getting destroyed? Will the number not explode once an 'A-OK' signal arrives from another star?

Attracting investors and obtaining their money by selling shares in the company is called equity financing. When the interstellar operation shoots out its first spaceship, it can perform an IPO and become a publicly traded company on the wave of news hype. As good news arrives, it issues shares and sells them at increasing prices. Debt funding, which is taking out loans for a bank, is not recommended here as the company does not have the cash to pay back interest. Equity investors, now called shareholders, can be asked to wait for the first dividends. 

Basically, the company can make money out of pure hype. This is similar to how tech start-ups operate. 

The second source of income, before any product has arrived, is data. Beyond instructions and reports, investors in the interstellar operation have no need of the data sent back by the colony. It costs nothing, however, to keep the dishes transmitting scientific and prospecting data. This data is sold to research bodies such as Universities for an additional revenue stream. The company might even convince a University to divert some of its astronomy budget towards helping the company along, in return for guarantees on delivering future data.

Another source of income is futures. These are agreement where the client agrees to buy a certain quantity of a certain product from the supplier at an agreed-upon price, in the future. The client pays now, and receives the products later. There are variations, such as the client allowing themselves to buy less than the expected amount, or the supplier offering prices that change with the future market, but the principle is the same: I get money now for something I'll do in the future.

In the case of an interstellar operation, investors can sell portions of their expected output to interested clients, and gain money in return. The clients will be happy that they'll get products at lower-than-market or even post-crash prices, while investors are happy to get money in return for something that cost them nothing to deliver. 

Futures are involved in every step of the operation, from selling and negotiating the price of products during the long journey back home, to trading in percentage outputs of the entire colony for the next 100 years. 

After the first delivery has arrived, there are many money-making opportunities for the operation's investors.
They can set up an artificial scarcity, where the massive amount of items delivered are sold slowly, over time, at higher prices than just dumping them in the market. They can offer specific items at the clients' demand in the next delivery, at a premium. Using their demonstrated success, they can ask for funding towards a second interstellar operation, headed towards another star system.

These will take the form of bonds. A certain amount of money is contracted towards the operation, to be returned at a future date. In the mean time, the interstellar company pays out interest. This interest is equal to the client's Required Rate of Return, and is called a coupon payment. 

Low returns on government bonds makes corporate bonds attractive.
Interstellar Bonds are necessarily long-term financial products. They match the 10 year treasury notes and 30 year TIPS (treasury inflation-protected securities) from the US government, or the long and ultra-long Gilts of the UK government. Unlike them, the interstellar operation can offer incredible coupon payments.

If multiple product deliveries succeed, the first clients that bought Interstellar Bonds will sell their contracts on the bond market, known as the secondary debt market. Successive high-profit events, such as the delivery of tons of rare material, are extremely useful for increasing the value of bonds. Once the bond price has reached a certain level, the operation can issue new bonds, at higher face values (contracted initial amounts) and lower coupon rates, and still be popular. A consistently well-performing bond market can become the company's primary source of revenue! 

The company can also offer indefinite bonds, which pay out a consistent 'rent', or short-term bonds just before a payload arrives. 

As mentioned above, the company is not limited to how much profit it makes off selling physical products. It can sell intangibles such as growth.

The Dividend Discount model for stock valuation. More growth means more valuable stock. 
For example, 10% of each payload's revenue is put towards an investment fund. That fund is used to increase the company's capacity for handling larger and larger payloads. By using a large array of lasers, the fraction of payload that can be sold is increased by 10 to 100 times or more compared to nuclear rockets. In the company's report, this is written down as 'growth is 10%'. It can then advertise a 10% increase in share value to potential investors. 

This can go even further.

Imagine it puts 1% of its profits towards a 'Second Interstellar Seed' fund. That 1% could open up an entire star system, with an entirely new stream of revenue. The actual value of the fund is dwarfed by its potential value... but what does the company sell? It sells potential value, of course.


Human travel to the stars is not going to be possible for a very long time due to the deltaV requirements. A small 'seed' of self-replicating components can cheat the rocket equation.

The seed produces a colony through exponential growth. A portion of the components are specialized into various units with specifc roles. Eventually, it can start exploiting local resources to send valuable products back to Earth.

Interstellar operations need the following three criteria to happen:

-A sufficient technological basis for fraction-C travel and reliable self-replicating machines (Realisable)

-A growth in resource consumption that increases costs to the point where interstellar alternatives are viable and/or (Profitable)
-An interplanetary presence and a Solar System closed to newcomers or legally or de-facto hostile to free-for-all investment makes such investments necessary (Affordable)

These conditions might take several centuries or more to come together. Having the technological ability to travel to other stars is not a guarantee of regular interstellar travel. Not having enough resources to continue growing will not push humanity to the stars, but into space, the asteroids and the moons of our Solar System.

The best method of considering interstellar travel is from a financial and economic standpoint. We do not invest in projects because we can or because we want to, but because they are better to the existing alternatives.

Money can be made from interstellar travel before a single kilo of precious metal is sold on the markets. Financial products can allow a company that invests in interstellar operations to produce a cash flow greater than just what is sold on the market, by selling intangibles such as rights, data, growth and investor confidence. By looking at interstellar operations as just another investment model that justifies the set-up cost with huge profit margins, we can realistically fit it into a setting where humanity has not yet left the solar system. 

Feel free to ask questions and discuss the topic.


  1. Really interesting article. I always thought shipping of bulk goods on an interstellar scale would be unprofitable, but this article convinced me I was just not looking far enough in the future. But aren't there a several articles on how exotic materials can often be replaced with less exotic element (but in a rare configuration, like graphene, or carbon nanotubes filed with zinc)? I think bulk shipping of rare goods will happen in the long run, as well as information (one of the few things that can travel at 1c), but it migh actually be slowed by advancing material sciences (really, with carbon fibre, diamond, carbon nanotubes, carbon nanotubes filled with a conducter, graphene, vitreous carbon, etc. I am surprised how versatile some very common elements like carbon or sulphur are).

    1. Thanks! I'm glad you liked it.

      The trick is to off-load the cost of interstellar travel to a colony that doesn't work on money.

      You are correct about materials sciences advancing towards using more common materials... but there are some properties that cannot be replicated. These include resistance to corrosion, catalytics, high temperature conductivity, special electronic properties such as germanium in lasers...

      Basically, the combination of rarity and usefulness means that such elements will always be in high demand.

  2. Can the AIs running things at the colony be kept from developing their own agendas so Sol system has to send things or data the AIs want to keep the goods coming?

    Jim Baerg

    1. I don't think any level of AI is needed beyond simple autonomous programs. If everything can be packed into the DNA encoding, then the colony knows how to do everything it needs to know how to do. No smart selection or reasoning or anything like that, just trigger certain reactions following critical points or data.

      For example, it can be told to eject a part of itself towards the nearest landmass as soon as it can afford to do so. This helps is spread throughout the solar system. It can deduct which are precious metals and what is useless ore through a chemical analysis. Higher concentrations of platinum for example could attract more mining components than other sites, helping it focus on higher value output.

      If the data is hard to fit in the seed, it can be sent later, like sending instructions of how to find Earth in the sky and how to shoot a big laser in that direction so that outbound payloads can hitch a ride.

      Over time, of course, the colony needs to be more intelligent. This helps it avoid traps like shooting sections of itself into a planet with an atmosphere and watching them burn up dumbly, again and again. To be discussed in Part 3, it will have to face 'smart' threats that try to divert its output or sabotage its output as a ransom for pirates back home.

      So yes, eventually it will need to level of intelligence of an AI... which is probably the point at which it might be meaningful to send a small detachment of humans to manage to system-spanning decade-old colony... that's actually a good idea. Send robot 'seeds' ahead, have them send back millions of tons of free equipment and materials over a century, then use all that free stuff to build fast spaceships to send actual humans to manage the huge colony for cheap!

  3. I can't speak to the rest, but the technology required to create a self-replicating bacterial colony that spits out starships is well beyond the weirdness event horizon where most stories simply cannot be told because there are no humans left to tell them.

    1. I don't quite agree. The 'starships' the bacterial colony would produce are not very complex. This topic will be covered in the next post, but the colony can afford to use very inefficient yet very simple solar concentrators and thermo-electric system to power huge, wasteful lasers as engines for returning payloads. These are the sort of things you can make in a 3D printer today.

      Even so, you can get away with 'dumb' colonies that receive blueprints and detailed instructions from home. The first thing the colony builds is a large radio receiver. It listens to instructions, and executes them until a new set is delivered.

      On Earth, the effect of such technologies is that bulk manufacturing would become massively cheaper. Instead of a large car factory, you'd have vats of programmable bacteria churning out simple items.

      Complex items, such as microprocessors and efficient lasers, would have to be built using expensive and complex 'regular' factories.

      Overall, a small effect.

    2. With respect, you don't know what you're talking about.

      If we have the technology to reprogram bacteria to produce any sort of laser*, no matter how primitive, then we already have all the technology necessary to engineer other organisms (and ourselves) in the true sense of the word.

      Which means that there won't be any humans around to run an interstellar colony as a long-term risk vehicle. There might be something else around instead, but it will be so alien that talking about joint stock companies or whatever is pointless.

      *An analogy: I show you a model rocket built with a simple black powder motor. You then point out that this rocket demonstrates all the principles needed for space travel and begin to speculate as to how many trips to mars we could make per month using a 'suitably scaled up' version. I should also note that I've worked with bacteria (simple stuff - plasmid construction and cloning in E. coli for use in plant transformation experiments) for what that's worth.

    3. To be fair, we are both speculating within the borders of plausibility and realism. I'd also like to point out that this blog tries to accommodate all sorts of settings, ranging from alternative history Cold Wars to ultra-futuristic space opera inspired by the Culture, so if programmed bacteria are unlikely or not depends mainly on the setting's author.

      But, back to the points you raised.

      If programmed bacteria, which already exist and are being developed towards producing drugs and fighting cancer, can be developed to the point where they can produce macro-scale structures such as refineries and lasers... it does not necessarily mean that we are able to engineer other organisms.

      A living, breathing artificial creature has dozens of macro-scale structures and literal thousands of type of micro-structures, from hormone glands to the little muscles which raise hairs when you're cold to trap air next to your skin. It is a much more complex task than building lasers!

      Also, creating artificial organisms is an entirely different path of research from programming bacteria to make machines, and it is unlikely that even the tools being used are similar.

      To finish on that topic, even if we can both make bacteria into miniature factories AND produce artificial organisms... how will this hurt the concept of sending tiny colony seeds to other start systems?

      Humans running a colony were never part of the concept described in this post - in fact, it would be very detrimental to interstellar trade because if humans are needed, then the minimum size of an interstellar operations becomes prohibitively expensive! No, it will be autonomous colonies that evolve based on the information received from us.

      I don't understand what you mean by 'something else around', sorry. Could you clarify?

    4. My point regarding bacteria is that any self-assembling machine is going to require some means of self-assembly. You could very easily make a bunch of bacterial strains which can act as components of a machine, but getting those components to grow into the machine itself puts you firmly within the range of engineering organisms from the ground up. Which bacteria aren't really suited for, but that's another conversation.

      Talking about sending a packet of cells over and having it grow into a factory is, compared to present-day biotech, the difference between chariots and modern car factories. And present-day biotech is already in danger of smashing the biological underpinnings of human society - including the 'human' part.

      My comment about 'something else' is that your civilisation capable of sending self-growing laser colonies to another star won't have anything we would recognise as human in it. It might have hundreds of different species of sapients floating around (some descended from us), or might have a single species of hyper-intelligent creatures whose thoughts we cannot fathom. It might not have anything we'd recognise as a mind at all - a vast radio-linked colony organism or a sentient ecosystem. We don't know.

      What it manifestly will not have is human actors as we understand them today. And that pretty much rules out any assumptions we can make to underpin concepts like long-term investments or joint stock companies.

    5. Bacteria are designed by nature to self-replicate. They do it by growing copies of their internal proteins and molecules, then dividing into two.

      The 'programmed' part is for them to produce unnatural proteins with artificial effects. For example, they could have specific triggers to start sifting through ingested matter for silicon. The silicon accumulated inside the bacteria, then it dies. The dead cell remains are eaten away by enzymes produced by another set of bacteria, leaving a deposit of silicon. A third set of bacteria produces acid. A fourth set deposits markers in specific patterns. The acid is used to etch lines and grids in the silicon. A fifth set deposits conductive metals.

      Eventually, an electrical circuit is produced. The circuit works like any inanimate machine - the bacteria just build it.

      I personally believe that developments in biotechnology that produce self-replicating colonies of programmed bacteria will be correlated with developments that allow humans to change themselves, but they are not necessarily causated. A new technology that allows bacteria to grow in patterns is not the same technology that grows new intelligent species. The techniques to build lasers using a group of cells is not the same technique for increasing intelligence in human brains. The tools will be similar, but one will not mean the other.

      And, even if all your statements are realized and we become a designer species before we start colonies in other star systems.... it does not invalidate the concept of sending small seeds over interstellar distances. That still stands.

      As for financial realities changing with the players, this is not true. Finance models and things like investments, stocks and interest rates are based on mathematics and will forever remain true. What people do with them or how much they trust or value them is a matter outside the scope of this blog post.

    6. Matter Beam, I would like to nitpick.

      As you said earlier: "A living, breathing artificial creature has dozens of macro-scale structures and literal thousands of type of micro-structures, from hormone glands to the little muscles which raise hairs when you're cold to trap air next to your skin. It is a much more complex task than building lasers!"

      Consider: Your bacterium colony would not only have to construct a laser and the cargo sails that ride it back, it would also have to construct the powerplant that powers the whole setup, mining gear to break up rock for biorefining, refineries for materials that can't be biorefined, transport mechanisms for everything-

      All this, from local materials, in an intelligent manner that best advances the objective of sending raw materials back, in one of the most inhospitable environments known to man.

      Your bacterium colony would thus seem to fit your description of an artificial organism, both in terms of makeup(scale differences aside) and difficulty of engineering.

    7. The level of complexity is still incomparable. Here is a general list of the number of different types of cells, systems and processes in the human body.


      About 200 types.


      79 different types of organ.


      10 systems.

      This ignores multiples of the organs, systems created from the interaction between organs in specific places (like saliva, tongue and mouth internal epithelial layer allowing us to pass food into the esophagus), as well as the literal billions of different types of chemical reactions, millions of drugs that can affect the body in different ways and the zottabytes per gram of data contained in DNA.

      The number of parts in a laser are much, much lower. Instructions for building laser are nowhere as dense as the information in DNA, and they're the most complex thing in the whole project... and even then, you can beam instructions afterwards without having to store them on the bacteria themselves.

      The whole point of programmable bacteria is to 'bolt-on' a certain level of artificial functioning to an organism far more complex than what can be achieved in a laboratory. Most of the hard work, such as converting sunlight into energy without any rare materials or replicating itself with a low error rate and no external infrastructure, is done by the natural part of the bacteria.

    8. Anonymous, I think our gracious host has his mind set on this concept.

      Argument from experience seems to have failed. Pointing out that his concept effectively is the design of a massive multicellular organism has failed. Pointing out that bacteria are capable of, at best, producing colonies rather than organisms (there's a reason that we're eukaryotes rather than bacteria) will similarly fail.

      Matter Beam, you are correct that there is nothing to stop some sort of biotechnologically advanced civilisation from sending over a seed over interstellar distances. But, again, this gets far into the weirdness boundary where the other capabilities of such a civilisation make it hard to determine whether the other assumptions will hold true. It may be that, with that level of technology at your disposal, you could simply send over a terraforming package and follow it up with a citizen-growing organism for populating your new colony. After all, why trade when you have the ability to simply export your civilisation wholesale? The underlying rationale driving the exercise, in other words, might be rendered moot.

      Regarding the idea that a bacterial colony capable of producing and shipping raw materials would be simpler than a eukaryotic organism complexity of structures and the data storage required - I simply disagree with you. That said: I'd love to hear any concrete proposals that you have for producing such devices, as these would be revolutionary even in a primitive form. Do you happen to have a scheme for the genetic pathways involved, the regulation of expression, a process flowchart or the like?

    9. Matter Beam, while Thom is making most of my argument, he has two specific assumptions we both share that he seems to only lightly touch on.

      Namingly, that

      *Bacterium cannot be expected to provide all or even most of the manufacturing processess required.
      *Creating and managing the requisite infrastructure for a complete and fully operational colony-that is, one capable of sending home goods as required while supporting itself and sustaining it's own growth- is substantionally more difficult than constructing and managing the laser array that actually sends the goods along.

      While I cannot comment on the first point due to lack of expertise in things biological, I will comment briefly on the second.
      Even assuming the best case scenario, that bacteria can be made to produce all the machinery required from raw materials in one step, the fact is that individual asteriods almost never have all the resources you need to expand and send home. C-types in particular, which you'll need for CHON to grow your bacteria, almost never have metals or silicon in substantial quality. You'd have to spread the colony to a few different asteriods to get everything you need.
      This requires interplanetary transport mechanisms to trade resources between asteriods. Also, radios and sosphisticated enough intelligences to coordinate and plan this activity.
      If the transport mechanism is some species of rocket, propellant needs must also be met. If it's momentum tethers, the momentum stored in each tether must be managed such that no one tether spends all of their momentum. Etc etc-there will always be a cost and management challenge associated with transporting stuff between nodes insystem, and one the colony have to solve in order to get anything done.

    10. *Quantity, blah.

      Thom, I do believe that last sentence was unneeded. Matter Beam is trying to provide a plausible mechanism for interstellar trade for fiction, so I think we can gloss over some of the finer details.

    11. Anonymous,

      Tone is hard to convey via text, so I should clarify: I would be genuinely interested in any reasonably comprehensive proposals as to a mechanism for getting your bacterial colony to manufacture objects, because it really would be a breakthrough worth investigating and commercialising. I'm sincerely hopeful that Matter Beam has inventive concept, as I believe that such things need not be conceptually complex, and need to be restricted to experts in the field.

      In terms of story, you are indeed allowed to fudge the details. The best SF, however, becomes stronger by thinking through the technology and its implications.

      Greg Bear wrote a very interesting biotech SF novel (blood music) which was made stronger by the fact that he had a semi-realistic concept for how his DNA computing was supposed to work.

    12. @Thom S:

      I'll have to be honest. The focus of this article was to place the energy cost and possible returns of an interstellar colony within a financially reasonable situation to allow it to become interesting sooner than later.

      From that resulted a list of requirements (interplanetary transport network, self-replicating colony).

      I tried to find existing concepts that matched those requirements and wrote about them. For example, the Laser Weapon Web can be converted into an Interplanetary Laser Web to handle boosting the payload to near-lightspeed, while the combination of bacterial manufacturing (http://www.creative-biolabs.com/Bacterial-Manufacturing.html), their ability to handle metals (https://en.wikipedia.org/wiki/Bioleaching) and a demonstrated ability to arrange themselves into artificial structures and perform machine-like tasks (https://link.springer.com/article/10.1007/s10404-007-0252-6) allow for self-replicating colonies.

      As noted, they are only one solution of several. Maybe we'll skip straight to dry inorganic nanotechnology before we get a good handle on bionanotechnology? I don't know! This is really the downside to ToughSF's approach to hard science fiction. Is there one good solution based on available data? Will the author's tastes (organic/inorganic solution) prevail due to the vagueness of the data available? I cannot cut one way or the other without disregarding the interests of authors who want to do things differently. I prefer to accommodate instead of alienate.

      Now to the question of complexity: I personally think that a laser is far less complex than a biological organ, that a colony of interacting systems can be done in an 'additional' manner than an 'integrated' manner and that managing the colony can be divided like it is done in our own body: blind sub-systems 'directed' by a superior system. In this case, the superior system is our investors beaming up specific instructions to the colony.

      Now I concede, my personal idea of the situation might have bled into the supposed objective assessment of the options available to authors for implementing the interstellar seed concept for trade between star system, but I must warn you against expanding the scope of the argument. Bionanotechnology on the level required to produce a self-replicating colony does have some serious implications for the rest of the setting, and I am happy to discuss them (the whole point of this blog!), but I don't want the secondary implications to affect the initial premise.

      For example, suppose we thought inorganic nanotechnology was THE best solution. There are clear examples of modern research and study into how nano-assemblers can perform the function of larger machines, which is perfect for extrapolating into how they could work into the interstellar seed concept. Would your arguments have been the same? I think not. Therefore, it is best to separate the discussion on bacteria-based bionanotechnology from the broader idea of interstellar trade and seed-colonies.

      Here, to restart that discussion: what if we used eukaryotic animal and plant cells instead of bacteria? Would some of the issues with complexity be solved?

    13. @Anonymous (find and use a name, even in signature!):

      The way I worked it out was that the bacteria were only the first layer of a larger, more conventional system of machines servicing each other.

      For example, the bacteria organically sift through a comet's volatiles and extract the resources they need to reproduce. The metals would be interacted with by specific, artificial enzymes and clustered in pockets where a second breed of bacteria would sort them by elements by chemical process, a third breed would eat and deposit select elements in specific shapes, a fourth breed layered the shapes, a fifth...

      ... and so on until we have the biological equivalent of a 3D-printed robot.

      The robot can now mechanically move ores around much more quickly than bacteria. The robot can build bigger versions of itself.

      The larger, mechanical 3D printer produce nozzles and solar panels. A solar thermal rocket can nudge around a comet into a rendezvous with a rocky asteroid.

      By the time we are producing lasers and returning products to the Solar System, the bacteria are a tiny portion of the larger colony. Their only function would be to become interplanetary seeds. An early colony would find it cheaper to shoot off a new seed to a nearby asteroid than to build a rocket to reach it. Afterwards, a mature colony will move around with spaceships and start extracting and refining using industrial machines, because the bacteria are too slow.

      As for the CHON availability and the choice of object to land on, consider this report: https://www.timeshighereducation.com/news/evidence-from-esas-rosetta-spacecraft-suggests-that-comets-are-more-icy-dirtball-than-dirty-snowball/199168.article

      Comets are likely to have a useful portion of their mass as 'dirt', and that dirt resembles the composition of planetary surfaces.

      Carbon monoxide, nitrogen all present: http://www.space.com/28884-rosetta-comet-nitrogen-discovery.html

      The Rosetta comet is 10000 million tons, and it isn't even one of the bigger comets in our solar system. If an interstellar seed lands on one of these, it has all the resources it could ever need to reach the later stages of colony development.

    14. Matter Beam,

      Thank you for the generous reply.

      My arguments regarding nanomachines would indeed be different, for the simple reason that I know nothing much about nanotechnology and its real-world possibilities. I'm accordingly much more naive and inclined to grant nanobots perhaps-magical powers than I am GMOs.

      Moving on: eukaryotic organisms (which include plants, animals and fungi) would actually work much better here, with the proviso that bacterial symbiotes might be part of the package as well (think nitrogen fixation and symbiotic cyanobacteria). Eukaryotes have essentially solved the issue of growing, differentiating and controlling a bunch of specialised cell types, so using their solutions would be a logical way to proceed. Chromosomes also have far less restrictive data storage capacities, so you would avoid the issue of your bacteria not having enough instruction space to perform the complex intracellular interactions needed to produce a designed structure.

    15. My take on a 'realistic' story which uses biotechnology but otherwise fits your scenario as closely as possible:

      Singularitarians are wrong - becoming smarter does not result in you being better-equiped to solve the generalised problem of intelligence. Instead, intelligence turns out to be a process which becomes exponentially more complex as you go, while providing only a linear return. We're proof of that already, after all - being smart (as in, high-IQ smart) seems to come with enough negative tradeoffs so that no fitness advantage accrued and human intelligence plateaued across the planetary population long before the invention of agriculture. So sapients aren't much smarter than they are now. They also aren't in danger of developing power sources or travel systems which completely break with the laws of physics as we know them. Fusion is still the best energy source available, even after thousands of years of development and refinement.

      Some sapients, however, are much more long-lived - hundreds of years in some carefully self-cultivated clades. And long lives allow for truly long-term investments.

      The seed to the stars is literally that - a small package of carefully designed organisms and starter nutrients sent on the long journey across the stars by a tiny but incredibly sophisticated ship. The ship, massing at most a few kilograms, is capable of delivering its cargo to a distant asteroid around a foreign star.

      Once delivered, the seed hatches into a variety of organisms which begin to produce a space-based ecosystem. They may be joined, at intervals, by more seeds containing different organisms for filling out more advanced niches created by the pioneer species.

      Over time (perhaps thousands of years) the initial ecosystem grows to encompass other solar bodies and becomes increasingly complex. The climax of this carefully-designed complexity is the creation and launch of return vessels bearing scarce resources back to the home system. Alternatively, the final series of seeds might be some sort of egg-ship for sapient clades rich enough to afford a trip for their offspring to the ultimate in luxury housing: an entire solar system turned into an Eden for them to live in.

    16. @Thom S:

      Well then there we have it! We'd need a biological package where bacteria mindlessly reproduce and create the appropriate environment for eukaryotic cells to start performing more complex tasks.

      I think the problem with quantifying 'nanotechnology' is that the term can be applied to such a broad spectrum of uses and capabilities. It might be as broad as grouping a dagger and an atom bomb under the category 'weapon'.

      The biggest problem with nanotechnology in these situations (the wild) is that they won't have comfortable, temperature regulated and top-down controlled environments to work in. They'd also have to gain a level of complexity unimaginable today to do their work without constant supervision. Just list the number of proteins and sugars dotting the surface of a human cell required to allow it to be transported to certain parts of the body, recognised as part of the body or to allow specific elements entry and exit from the cell. A nanobot would need similar systems to prevent it by being consumed by its neighbour, continuing to function after being damaged or to signal that it needs more energy ect.

      This round-about argument is to make the point that relying on the natural solutions to all these problems and just piggybacking our rough modifications on living cells is much easier to accomplish. By the time we can recreate the majority of the functions of a self-sufficient cell in an inorganic format, we might as well use the cell.

    17. @Thom S:

      Certainly an interesting assessment of intelligence: there's a lot you can write from that!
      However, are you sure that a colony on another star system would take "thousands of years" to become usable? Unless your target is the terraformation of an entire planet, then the colony will likely become profitable within a few years. Exponential growth with uncapped access to energy and raw materials will allow any self-replicating colony to expand at close to the maximum rate of growth... this is an exponential growth.

      Also, the transport system that returns payloads to the home system can be used in reverse: to capture a large spaceship without requiring a second stage.

      This means, theoretically, as soon as you start receiving 1 ton or larger products, then you can return 1 ton or larger spaceships and expect them to be captured into the destination system...

    18. My uninformed suspicion about nanotechnology is that it ends up looking a lot like biotechnology. But that ruins all of the wonderful 'nanotech as magic' stories that power so much modern SF.

      In terms of time span, I am being conservative as I don't know what the bottleneck in a cometary ecosystem would be. If the bottleneck is something like a micronutrient, then you might see fairly rapid colonisation and ecological succession.

      My concept for how the initial colonisation would work is that your keystone organism would be something like a space-adapted lichen or algae; which would cover the surface of the asteroid in a thick mat of biofilm. The mat would insulate the surface - allowing liquid water to form underneath and preventing outgassing to an extent. Secondary organisms would work to extend the liquid layer and extract nutrients. Some would live entirely under the mat, while some would poke structures through into space in order to harvest sunlight. The asteroid would become a miniature jungle floating on an expanding sea.

      In time, the climax organisms which provide the actual goods would emerge. These would probably be animals of some sort; capable of feeding on the biomass stored in the mat and energy harvesters.

    19. Matter Beam,

      I went away, came back, and reviewed this topic again, and had a thought.

      Part of the problem is the management and computational aspect-the colony would need to decide how to allocate resources so as to best advance the mission on it's own, as Mission Control have years of communication lag to deal with. In addition, several tasks require calculation more sophisticated

      DNA computing might be useful in this. We would use the existing molecules in the organism to perform computations without dedicated neuron cells. We only really need fast reaction times for maneuvering in space near other objects, such as docking a material transport to a colony or fire control for the laser array, which can be handled by conventional electronics or dedicated neurons.

      This would also mean that the organism itself can be made programmable, without hard-coding all the instructions into the DNA itself-just enough to set up the 'computer.'

      As an aside, it would be nice for you to cover DNA computing sometime-perhaps as a sequel to your nanotechnology post.

      P.S. Got a name.

    20. "We only really need fast reaction times for maneuvering in space near other objects, such as docking a material transport to a colony or fire control for the laser array."
      Should be
      "We only really need fast reaction times for things like as docking a material transport to a colony or fire control for the laser array,"
      Too bad I can't seem to edit my comment.

    21. DNA computing is certainly a solution to many of the colony's problems, but current concepts have a major fault: sorting through and reading the data produced by the molecules requires a computer nearly as powerful as the DNA computer itself!

      One way around this problem is to simply use the DNA as a storage medium that allows simple cells to build more efficient electric brains, and those brains 'read' software out of the DNA to become reasonably effective management programs.

      Another problem with all sorts of chemical computing is that they work best when given a lot of options, and have to test each one. Parallel processing is excellent for defeating combinatorial problems. Managing a colony is more of a problem of identifying opportunities, matching situations to problem-solving methods and optimizing responses, such as, not trying to do the same thing over and over when the first time failed. DNA computing would not help much in this case...

      I'll add 'Future Computing' to my list of topics, thanks!

    22. "Another problem with all sorts of chemical computing is that they work best when given a lot of options, and have to test each one."

      This seems like half the required for managing the colony-simulating possible courses and futures.

      The other would be paring down generated solutions until you have just the best possible ones. Maybe have the DNA computer output solutions with a molecule that carries a variable for desirability, and then have an enzyme that grabs two solutions and cuts up the solution with less desirability?

      Just a random thought.

    23. I'm not too well versed in computing, but I do know that project management has too many 'desirability' factors and intractable issues for a powerful yet simplistic tool like chemical computing to solve.

      The hardest part would be converting the complicated multi-variable management problem into a mathematical problem that a computer can solve. This is a problem even today and some very strong AI would be necessary to perform this task autonomously. Then, once you've got your mathematical model, you'll realize that it is only valid for that specific situation and that number of variables. If you encounter a 'pop-up threat', you'll have to devise a new one.

      Either way, project management greatly reduces the utility of having super-fast computing, as most of the time and resources would be dedicated to understanding the situation and figuring out how to convert it into solvable models anyway.

    24. Why not close the loop? If you're already going to use biology to seed self replicators to do difficult tasks, grow a biological AI onsite once you need one. Don't send people, send their genes! One of your first things you'll want is a device that transforms information from the corporate overlords into an arbitrary sequence of genes. This level of DNA writing technology already exists. Use that to assemble human gametes (this technology does not exist) and grow people onsite. This approach can generally cut mass requirements for any biotech you want to send. Or just include some frozen human embryos in your seed package. The first generation will come out pretty weird being raised by AI systems and cometary jungle lizards, but they'll provide your strong AI without ponying up the tremendous expense of sending whole people.

      There is also the issue of indoctrinating them to put the company first. Humans are not the most loyal and reliable of systems and these isolated people could conceivably stop shipping things until their demands are met. They will likely have a very different perspective raised by bots, literally grown to command an ever expanding industrial complex the size of a solar system. Us Sol-grubbers might find them...alien. Humanoid aliens! Don't piss them off, that same giant industrial complex is literally designed to fire as many r-bombs as possible in as little time as possible.

      This is all assuming no uploading, since that makes travel just information flow.

    25. Encoding enough information to be executed by self-replicating cells to make something functional like a radio dish to receive further instruction is going to be wildly difficult. Doing the same to 'grow' an AI out of a few grams of radiation hardened soup is even more difficult than that!

      You probably don't want to send human DNA itself. You get your self-replicators to build a suitable environment and give them the ability to create complex bio-molecules out of local materials.

      I don't really think the first generation of humans will grow up to be very weird. Machines can emulate humans pretty well for the sake of your infants, and as they grow up they'll have each other and a planet's worth on data on how to keep the kids sane.

      How they turn out morally is... an entirely different question. There's some good scifi material in there.

    26. However you build the dish,I was merely pointing out that a sequencer building radio transmitted plans is a big mass advantage. Whatever combination of comet soup bacteria and robots you're using, you're still building a radio dish.

      The biological AI I'm referring to is the humans themselves. I'm not so sure how well we can build nanny bots, but I'm pretty sure whatever the build looks like it'll make pretty different kids than a culture saturated with people. Perhaps not inhuman, but they'll have a very different experience growing up with only a few peers to talk to compared to the uncounted billions teeming around Sol.

    27. Really enjoying this blog! I've been into hard scifi worldbuilding for a long time now, so this is right up my alley.

  4. I find your article very helpful, because I am designing my fictious universe with STL interstellar travel :3
    But, in my universe, the distances between stars are quite small in a few lightmonths to 2 lightyears range

    1. That's an ingenious solution to interstellar travel: just put everything close to each other.

    2. If the stars are close enough to have significant influence over each other, enough to orbit a center of gravity between them and cause tidal effects on the planets around each, then you could have cases of planets being flung out of the star systems.

      They'd make for interesting destinations.

  5. Colonies that can either grow themselves or attend to non-growing but productive tasks.....

    Sounds like the old 'invest interest in capital or spend it' argument.

    'My next question sir, is whether you want to let the colony expand further with minimal productive activity. In this business we call that 'reinvesting the interest' and it is the lowest risk strategy. On the other hand, you could have it produce goods now for return to earth with minimal self-expansion. That is a higher risk strategy of course, but isn't out of the question.'

    1. The growth rate and specialisation fraction would be determined by models trying to find the optimal output vs growth ratio. The decisions to be made will be discussed in Part 3.

      One modern financial caveat to 'reinvesting the growth' is that the share price integrates the future value of a growing investment. Selling these shares allows you to cash in on your growth without increasing your output.

  6. If I'm correctly understanding the points being made here, the main emphasis is having products or commodity's that are profitable enough, and interesting enough to the business communities to commit vast amounts of resources to be tied up for very long amounts of time, with the promise of really big returns at the end. One such possible scenario I can see is one much like the current movie 'Passengers', where a group of very rich people pay through the nose to be boarded in stasis on a star ship, travelling to a new colony exo-planet for a much better life. Another possibility may be that a way has been found to ram scoop anti-matter, dark matter, or exotic matter on some sort of super tanker ship in interstellar space in quantities impossible on Earth. But only if done while travelling very far away at insanely fast speeds. Or perhaps an unduplicate-able substance found only on a far away planet that can extend human life, restore youth, or cure an otherwise incurable disease. Any of those can make a decent plot in a sci-fi where such technology is available.

    1. The things you mention are usually called 'McGuffinite': a substance or service that can only be obtained in space, and is sufficient reason to go there.

      Interstellar Trade Is Possible focuses on a realistic method for trading with the stars without requiring McGuffinite. Using simple financial analysis, we can determine that a self-replicating colony that grows indefinitely will always become cheaper to operate than an interplanetary version that can only encroach onto human space.

      The use of micro-scale spacecraft reduces the 'vast amount of resources' to a much more manageable version.

      So, the emphasis is not on the value of the products, but the simple fact that interstellar products will always end up being competitive.

  7. I need a little mathematical help from someone who know calculus. (a subject I now wish I had taken at some point in my past). If an advanced propulsion space ship was accelerating at 1 G. from zero, how fast would it be going after half a billion kilometers? Would such speeds double, and triple if accelerating 2 G's, or 3 G's? or is that an entirely different formula? Finally how far at 1 G acceleration would it take to reach 50% c?

    1. Ah, the acceleration-distance formula.

      We know that Velocity = Acceleration * Time.
      We know that Distance = 0.5 * Acceleration * Time * Time

      1G to 0.5 billion km.
      9.81m/s^2 to 5*10^8m.
      5*10^8 = 0.5 * 9.81 * T * T
      T * T = 101936799
      T = 10096
      V = 10096 * 9.81 = 99045
      The spaceship will reach 99km/s.

      1G to 50%c
      9.81m/s^2 to 1.5*10^8m/s.
      V = A * T
      T = V/A = 1.5*10^8/9.81 = 15290519
      D = 0.5 * A * T * T = 0.5 * 9.81 * 15290519 ^ 2
      D = 1146788990825688 = 1.15 trillion km

  8. Question for the economy within the solar system: What industries/ trades/ locations would prefer a traditional rocket over a laser-launched lightcraft?

    1. To clarify: I mean when travelling between planets- nothing to do with chemical rockets from surface to orbit.

    2. Well, laser-launch runs on money and energy.
      We cant exactly tell right now which places will be able to afford it in the future, but we can discuss energy.

      Mercury will definitely favour beamed propulsion. It would be silly not to use the abundant sunlight.

      Stations orbiting Venus would get 175% of the expected sunlight due to the reflective cloud layer, so they would be inclined to developing the system. Aerobraking in and riding a laser out would be very cheap.

      Its hard to say what a future Earth would do. Sunlight is still plentiful, but the outrage of installing big honking weaponizable lasers over the heads of population centres would have to balanced against the desire not to smog up low orbits with radioactive exhaust trails and tons of fissile fuel sitting in reactors. I think power beamed up from the ground through an orbital relay would be very likely.

      Jupiter has its natural electrical power station Io, making everything cheap. Lasers would be used there for sure.

      Everywhere else is more likely to use traditional 'independent' rockets. Asteroid hoppers are unlikely to trail a gigantic laser behind them. Saturn and Uranus do not have the benefit of plentiful sunlight, so will rely on nuclear power. Using nuclear power directly in a rocket engine instead of converting it into electricity, beaming it, absorbing the laser and converting the power back into heat and propulsion will be much more efficient and give you more deltaV per kilogram of fissile fuel.

      I hope this answers some of your question.

    3. It does thankyou!

    4. There is the matter of scaling things up. A large laser battery orbiting Mercury might become more feasible/economical if the cost can be amortized by powering ships across the solar system.

      Since Robert L forward had suggested Terawatt laser systems orbiting Mercury could power starships (using massive fresnel lens systems scattered across the solar system), there seems no reason such a system could not be able to accommodate spacecraft moving from Neptune to Uranus, for example. Indeed, the Mercurials could drive most other competitors out of business by building more laser transmitters and spreading more beam aiming and focusing devices across the solar system faster than their competitors, and being where sunlight is most abundant, have dramatic cost advantages as well. This is very much like the laser web concept that was discussed in one of the early posts in this blog.

      Indeed, the only reason to have "self propelled" ships would be for the equivalent of the Coast Guard or Navy, where not being dependent on the power beam would allow for independent actions. Even there, CG and Naval vessels would probably use the power web for economy, and because that is where most of the possible targets or vessels in need of investigation or saving would be.

    5. @Thucydides: Bearing in mind that a self-propelled craft can go on its way even when there's a long queue to use the laser network. I'd be more convinced by a 'mixed economy' involving self-propelled craft and Mercurial lightcraft working together rather than one pushing out the other- especially when demand for the network is high.

    6. @Thucydides:

      I make this point in Part 2, which deals with transport energy and cost, that trying to write off the investment in a massive laser battery by saying it will sell electricity to Earth on the downtime is a silly idea. Robert L Forward is a bit too naive in that regard.

      The Laser Weapon Web, under its Laser Transport Network guise, will be explored in detail as well. There are also the kinetic streams and particle beam alternatives that have to be mentioned as viable alternatives if large scale laser technology doesn't become as cheap as we hope for.

      Small nitpick on the final point you made - laser thermal technology, in the cheapest continuous wave version, is very likely going to be less efficient than the on-board nuclear rockets that military spacecraft will use. They provide power, yes, but not propellant. So a high Isp nuclear electric or gas-core rocket will want to stick to its own propulsion system that try to use 1000s Isp max laser thermal systems.

      In fact, if we really go for cost effectiveness, laser thermal Isp will be around 400-500s using water or carbon dioxide. The can be packed into extreme mass ratio propellant tanks, and do not cost energy to produce like liquid hydrogen/liquid oxygen splitting and condensing processes. This makes them competitive compared to chemical fuels (slightly higher Isp AND cheaper propellant) and nuclear rockets (no on-board reactor).

      The table turn if the laser transport network has a 'premium' service composed of high-energy pulsed lasers that can ablate solid propellants to achieve both incredible exhaust velocities and stupid power-to-weight ratios...

    7. @Geoffrey S H:

      There is definitely more to be said about interplanetary laser webs and how they could plausibly be built up in a setting where both independent (nuclear) and alternative (kinetic stream, particle beam) propulsion is available.

      However, to keep within the spirit of this blog, I cannot start imposing specific assumptions on the settings the readers might have for the concepts described to remain valid. I believe this was a failing of Rocketpunk Manifesto, where many points raised and solutions described were specific to Rick's vision of a slightly anachronistic, sometimes contrived future dubbed the 'Plausible Midfuture'.

    8. I can see how different starting assumptions make for different outcomes. The main issue that I was thinking about is economics. Even with an on board nuclear power plant, tapping into a "laser web" allows for greater economy, although it also ties you to the web.

      Military or "Space Guard" craft will likely still want to or need to use the web, partly because that is where most of the space traffic will be, and partly to economize on the use of the expensive nuclear power plants. Once they come into range of the target, it is easy to cast off from the Web and fire up the nuclear engine, with the added bonus that you are not needing to use any of your on board energy or reaction mass to get up to velocity, all your deltaV cam be used for combat or rescue manoeuvres.

      Given the very different operating environments, the nuclear powered spacecraft will likely be sitting on a boost or trans stage which used the laser web for power. How it gets back to the laser stage could be interesting; use too much of your nuclear reactor remass and you might not be able to dock and ride the laser beam home.....

    9. @Thucydides: Quite true. It's up to the author, but the point of this blog, a always, is to give you the tools to make a diverse and interesting setting grounded in reality.

      For example, I can easily see a setting where nuclear materials are government controlled. The solid cores in NTRs are 'loaned' and you have to submit documentation and detailed flight plans to use them at all. This would be the essential distinction between military/space guard craft and civilian ones.

      Another thought is that chemical rockets become useful for combat and other off-grid operations if a laser booster is involved. Even more so if the lasers are of the ultra-high Isp pulsed type. 5-10km/s deltaV is easily achieved with a set of drop tanks, and the thrust generated rivals or exceed nuclear engines...

  9. Quite interesting article!

    P.S. Assuming, of course, that there would be still more-or-less modern-type capitalistic economy by the time when interstellar trade would became technically possible...

    1. Ah well, we all need some sort of anchor point to keep the fiction relatable to the audience. Its also very hard to judge whether you are working out the consequences correctly without some sort of ceteris paribus: if everything changes at once (the technology, the economy and the people), you can't explain how we moved from our current situation to the potential setting.

    2. Markets seem to be "natural" in the way ecologies are, an efficient mechanism for ordering flows of energy, information and capital. "Capitalism" is the currently best efficient known way of tapping into and utilizing markets, and market capitalism has existed for thousands of years (ancient Phoenician merchants would not have much difficulty understanding today's market, they would be amazed more at the scale and scope of things, and the sheer speed of market transactions).

      Needless to say, if there is some sort of post capitalist mechanism to tap markets, it is by definition beyond our understanding at this time (since no one has invented it yet).

  10. With respect to the discussion about claims in space, some think the solution is to establish a legal code, for Mars at least, now: marslegalcode.org.

  11. Sorry to pop this question in late, but how does fuel/transport energy cost figure into this? Is the price of deuturium and antimatter (and containment costs) and it's production figured into the transport cost of goods? How much is an interstellar pack of Twinkies going to cost me?


    1. Hi!

      In the traditional shipping method, you load up a transport ship with fuel and have it travel out to its destination, load up on goods and come back. The cost of the fuel and other running costs have to be compared to the value of the goods, hence your question.

      In this shipping scheme, you send out one small ship with a factory. The factor builds 100, 1000, 1000000 or more ships loaded with cargo that come back on their own over a long period of time. The traditional comparison does not hold up. In other words, the value of the initial investment should always be significantly lower than the value of the products returning autonomously.

  12. CO2 + UV laser = C + O2... 3d bioprinting = Immortality = go to stars ((typewrite: interstellar travel constant acceleration))

  13. ...interstellar travel constant acceleration (Chiquitita the stars are shining for you over there in the high)… Today there is not any song, crying bitter tears for the Chiquitita from Mexico who never more will return to sing, she was shut up forever of the more inhuman mode by the coward feminicidal "machismo", in reality those coward assassins are not males, they are homovices who although have the strength of men, they are not men (only the homovices, TV propagation, hate to women, libro homovicio "La últ muj de Austrl", the last Australia´s woman, the homovices want make a world without women for "substitute to women", damned-homovices-dirty-repugnant-wretched-sterile-ill-vicious-mistaken). Politicians hypocrite politicians continuously doing damn´s concentrations to the coward feminicides perpetrated by the homovices, that "damn" DON´T IS USEFUL FOR NOTHING, as can see, the unique that is useful is the FEAR of Death Penalty, and all other are tales. The "civilized" politicians, guided by the religious Inquisition all credos "leaders" of the horror, they prefer that assassinating three thousand five hundred millions of women before call to Referendum and that being the People who decides applying to the feminicidal, a road-roller starting by the feet, only have that announce it and finished the cowards feminicides committed by the homovices. The horror against women will stop someday. Chiquitita las estrellas brillan por ti allá en lo alto.

  14. (2a)...interstellar travel not acceleration constant (laser)… a light´s beam that rotates is not a light´s beam that rotates… if it would be 1 only beam even the infinitesimal centrifugal force from gyration perhaps would break the beam launching tangentially its photons at straight line, as the stone of a sling loop that breaks. The arriving light from a gyratory pulsar star is something seemed, the same that gyratory laser pointer from horizon to horizon 180º in 1 second, each beam that exits from the gyratory source is an independent light´s ray with limited length that only goes away in straight line: source on, beam starts; source turns, that beam finishes, is off, and starts-on another beam in the new direction… Photon, its mass "is believed" that is zero, but… Photon, "relativistic" Mass M = E/c²... constant h=6.626*10^-34 joules-sec... v, frequency e.g. red light=4*10^14 Hz...that photon energy E = h*v; E = (6.626*10^-34) * (4*10^14); E = 2.6504*10^-19...that photon Mass M = E/c²; M = (2.6504*10^-19) / (9*10^16); M = 2.944889*10^-36 kgs… that Photon "relativistic" Mass ~29 ten-sextillionth of kg ______ 4 Screens, infinitesimal Centrifugal Force in kgs from that Photon, according to huge centrifugal G... 1: 95493 kms (96,105,971g)=2.8*10^-28 kgs (~28 hundred-thousand- quadrillionth of kg)… 2: Moon 384,403 kms (386,870,541g)=1.1*10^-27 kgs (~11 ten-thousand-quadrillionth of kg)… 3: Sun, 1.5*10^8 kms (150,962,873,002g)=4.4*10^-25 kgs (~44 hundred-quadrillionth of kg)… 4: Andromeda, 1.9*10^19 kms (913,135,316,143,368g)=2.7*10^-21 kgs (~27 ten-thousand-trillionth of kg)… When source turns, each individual laser mark on the screen has zero speed, do not moves, arrives and it vanishes, such as a light-bulbs row that they go being on and off one after another, from the first to the last hyperluminal speed, but is Not the speed of a mobile because there is Not any mobile.

  15. I didn't see this mentioned, so I was wondering how the seed package would decelerate at its destination, especially if it is laser-propelled?

    1. A magnetic sail can first brake against the ambient plasma of the interstellar medium, then brake against the solar wind of the target star.

    2. Thanks for explaining.

  16. This may be a stupid question, but : the whole premise of the operation would be to use an expensive, extensive infrastructure (likely beamed propulsion from a system-wide web or a trackway as described in the last part) to shoot the seed at 0.2-0.3C to the concerned economic zone for self-replication and the like; just so, the seed will use local energy and resources to build a propulsion system to shoot objects back to the home system. But, couldn’t the people of the home system itself order a self-replicating, assembling ‘seed’ to build a similar propulsion system to shoot the ‘actual’ seeds to the desired locations; this would forego any need to rely on prebuilt, extensive infrastructure for launching, right? Or is there something I’m not understanding?

    1. It is reasonable to assume that extensive interplanetary transport infrastructure would be established before sophisticated self-replicating 'seed' probes are developed. So using that infrastructure to send off the first probe is a logical cost-saving measure. It's not some strict requirement to set up interstellar trade!

    2. Thanks for relying; one last question : if the seed probe can go on to eventually design propulsion systems that are capable of massive energy outputs (transporting millions of tons of material back to the home system, albeit very inefficiently) by simply using the power of the nearby star, couldn’t the home civilisation also simply design a seed probe to grow and eventually use the energy of their home star to virtually supply them with limitless, ‘free’ energy (minus the operating cost for the probe, which should be trivial) for their own use?

    3. Yes, that would be possible!

  17. Thanks for clarifying