So, I'm going to talk about Elon Musk again, everybody's least favourite eccentric billionaire asshole and poster child for the Thomas Edison effect—get out in front of a bunch of faceless, hard-working engineers and wave that orchestra conductor's baton, while providing direction. Because I think he may be on course to become a multi-trillionaire—and it has nothing to do with cryptocurrency, NFTs, or colonizing Mars.

This we know: Musk has goals (some of them risible, some of them much more pragmatic), and within the limits of his world-view—I'm pretty sure he grew up reading the same right-wing near-future American SF yarns as me—he's fairly predictable. Reportedly he sat down some time around 2000 and made a list of the challenges facing humanity within his anticipated lifetime: roll out solar power, get cars off gasoline, colonize Mars, it's all there. Emperor of Mars is merely his most-publicized, most outrageous end goal. Everything then feeds into achieving the means to get there. But there are lots of sunk costs to pay for: getting to Mars ain't cheap, and he can't count on a government paying his bills (well, not every time). So each step needs to cover its costs.

What will pay for Starship, the mammoth actually-getting-ready-to-fly vehicle that was originally called the "Mars Colony Transporter"?

Starship is gargantuan. Fully fuelled on the pad it will weigh 5000 tons. In fully reusable mode it can put 100-150 tons of cargo into orbit—significantly more than a Saturn V or an Energiya, previously the largest launchers ever built. In expendable mode it can lift 250 tons, more than half the mass of the ISS, which was assembled over 20 years from a seemingly endless series of launches of 10-20 ton modules.

Seemingly even crazier, the Starship system is designed for one hour flight turnaround times, comparable to a refueling stop for a long-haul airliner. The mechazilla tower designed to catch descending stages in the last moments of flight and re-stack them on the pad is quite without precedent in the space sector, and yet they're prototyping the thing. Why would you even do that? Well,it makes no sense if you're still thinking of this in traditional space launch terms, so let's stop doing that. Instead it seems to me that SpaceX are trying to achieve something unprecedented with Starship. If it works ...

There are no commercial payloads that require a launcher in the 100 ton class, and precious few science missions. Currently the only clear-cut mission is Starship HLS, which NASA are drooling for—a derivative of Starship optimized for transporting cargo and crew to the Moon. (It loses the aerodynamic fins and the heat shield, because it's not coming back to Earth: it gets other modifications to turn it into a Moon truck with a payload in the 100-200 ton range, which is what you need if you're serious about running a Moon base on the scale of McMurdo station.)

Musk has trailed using early Starship flights to lift Starlink clusters—upgrading from the 60 satellites a Falcon 9 can deliver to something over 200 in one shot. But that's a very limited market.

So what could pay for Starship, and furthermore require a launch vehicle on that scale, and demand as many flights as Falcon 9 got from Starlink?

Well, let's look at the way Starlink synergizes with Musk's other businesses. (Bear in mind it's still in the beta-test stage of roll-out.) Obviously cheap wireless internet with low latency everywhere is a desirable goal: people will pay for it. But it's not obvious that enough people can afford a Starlink terminal for themselves. What's paying for Starlink? As Robert X. Cringely points out, Starlink is subsidized by the FCC—cablecos like Comcast can hand Starlink terminals to customers in remote areas in order to meet rural broadband service obligations that enable them to claim huge subsidies from the FCC: in return they get to milk the wallets of their much easier-to-reach urban/suburban customers. This covers the roll-out cost of Starlink, before Musk starts marketing it outside the USA.

So. What kind of vertically integrated business synergy could Musk be planning to exploit to cover the roll-out costs of Starship?

Musk owns Tesla Energy. And I think he's going to turn a profit on Starship by using it to launch Space based solar power satellites. By my back of the envelope calculation, a Starship can put roughly 5-10MW of space-rate photovoltaic cells into orbit in one shot. ROSA—Roll Out Solar Arrays now installed on the ISS are ridiculously light by historic standards, and flexible: they can be rolled up for launch, then unrolled on orbit. Current ROSA panels have a mass of 325kg and three pairs provide 120kW of power to the ISS: 2 tonnes for 120KW suggests that a 100 tonne Starship payload could produce 6MW using current generation panels, and I suspect a lot of that weight is structural overhead. The PV material used in ROSA reportedly weighs a mere 50 grams per square metre, comparable to lightweight laser printer paper, so a payload of pure PV material could have an area of up to 20 million square metres. At 100 watts of usable sunlight per square metre at Earth's orbit, that translates to 2GW. So Starship is definitely getting into the payload ball-park we'd need to make orbital SBSP stations practical. 1970s proposals foundered on the costs of the Space Shuttle, which was billed as offering $300/lb launch costs (a sad and pathetic joke), but Musk is selling Starship as a $2M/launch system, which works out at $20/kg.

So: disruptive launch system meets disruptive power technology, and if Tesla Energy isn't currently brainstorming how to build lightweight space-rated PV sheeting in gigawatt-up quantities I'll eat my hat.

Musk isn't the only person in this business. China is planning a 1 megawatt pilot orbital power station for 2030, increasing capacity to 1GW by 2049. Entirely coincidentally, I'm sure, the giant Long March 9 heavy launcher is due for test flights in 2030: ostensibly to support a Chinese crewed Lunar expedition, but I'm sure if you're going to build SBSP stations in bulk and the USA refuses to cooperate with you in space, having your own Starship clone would be handy.

I suspect if Musk uses Tesla Energy to push SBPS (launched via Starship) he will find a way to use his massive PV capacity to sell carbon offsets to his competitors. (Starship is designed to run on a fuel cycle that uses synthetic fuels—essential for Mars—that can be manufactured from carbon dioxide and water, if you add enough sunlight. Right now it burns fossil methane, but an early demonstration of the capability of SBPS would be using it to generate renewable fuel for its own launch system.)

Globally, we use roughly 18TW of power on a 24x7 basis. SBPS's big promise is that, unlike ground-based solar, the PV panels are in constant sunlight: there's no night when you're far enough out from the planetary surface. So it can provide base load power, just like nuclear or coal, only without the carbon emissions or long-lived waste products.

Assuming a roughly 70% transmission loss from orbit (beaming power by microwave to rectenna farms on Earth is inherently lossy) we would need roughly 60TW of PV panels in space. Which is 60,000 GW of panels, at roughly 1 km^2 per GW. With maximum optimism that looks like somewhere in the range of 3000-60,000 Starship launches, at $2M/flight is $6Bn to $120Bn ... which, over a period of years to decades, is chicken feed compared to the profit to be made by disrupting the 95% of the fossil fuel industry that just burns the stuff for energy. The cost of manufacturing the PV cells is another matter, but again: ground-based solar is already cheaper to install than shoveling coal into existing power stations, and in orbit it produces four times as much electricity per unit area.

Is Musk going to become a trillionaire? I don't know. He may fall flat on his face: he may not pick up the gold brick that his synergized businesses have placed at his feet: any number of other things could go wrong. I find the fact that other groups—notably the Chinese government—are also going this way, albeit much more slowly and timidly than I'm suggesting, is interesting. But even if Musk doesn't go there, someone is going to get SBPS working by 2030-2040, and in 2060 people will be scratching their heads and wondering why we ever bothered burning all that oil. But most likely Musk has noticed that this is a scheme that would make him unearthly shitpiles of money (the global energy sector in 2014 had revenue of $8Tn) and demand the thousands of Starship flights it will take to turn reusable orbital heavy lift into the sort of industry in its own right that it needs to be before you can start talking about building a city on Mars.

Exponentials, as COVID19 has reminded us, have an eerie quality to them. I think a 1MW SBPS by 2030 is highly likely, if not inevitable, given Starship's lift capacity. But we won't have a 1GW SBPS by 2049: we'll blow through that target by 2035, have a 1TW cluster that lights up the night sky by 2040, and by 2050 we may have ended use of non-synthetic fossil fuels.

If this sounds far-fetched, remember that back in 2011, SpaceX was a young upstart launch company. In 2010 they began flying Dragon capsule test articles: in 2011 they started experimenting with soft-landing first stage boosters. In the decade since then, they've grabbed 50% of the planetary launch market, launched the world's largest comsat cluster (still expanding), begun flying astronauts to the ISS for NASA, and demonstrated reliable soft-landing and re-flight of boosters. They're very close to overtaking the Space Shuttle in terms of reusability: no shuttle flew more than 30 times and SpaceX lately announced that their 10 flight target for Falcon 9 was just a goalpost (which they've already passed). If you look at their past decade, then a forward projection gets you more of the same, on a vastly larger scale, as I've described.

Who loses?

Well, there will be light pollution and the ground-based astronomers will be spitting blood. But in a choice between "keep the astronomers happy" and "climate oopsie, we all die", the astronomers lose. Most likely the existence of $20/kg launch systems will facilitate a new era of space-based astronomy: this is the wrong decade to be raising funds to build something like ELT, only bigger.

Image by Christopher Michel, via Wikimedia Commons

I am a big admirer of Kevin Kelly for the same reason I am of Brian Eno—he is constantly thinking. That thirst for knowledge and endless curiosity has always been the backbone to their particular art forms. For Eno it’s music, but for Kelly it’s in his editorship of the Whole Earth Review and then Wired magazine, providing a space for big ideas to reach the widest audience. (He’s also the reason one of my bucket lists is the Nakasendo, after seeing his photo essay on it.)

On his 68th birthday in 2020, Kelly posted on his blog a list of 68 “Unsolicited Bits of Advice.” One bit of advice that frames his thought process and his work is this one:

“I’m positive that in 100 years much of what I take to be true today will be proved to be wrong, maybe even embarrassingly wrong, and I try really hard to identify what it is that I am wrong about today.”

However, the list is more about wisdom from a life well-spent. Many fall into the art of being a curious human among other humans:

• Everyone is shy. Other people are waiting for you to introduce yourself to them, they are waiting for you to send them an email, they are waiting for you to ask them on a date. Go ahead.
• The more you are interested in others, the more interesting they find you. To be interesting, be interested.
• Being able to listen well is a superpower. While listening to someone you love keep asking them “Is there more?”, until there is no more.

And this is probably the hardest piece of advice these days:

• Learn how to learn from those you disagree with, or even offend you. See if you can find the truth in what they believe.

Other bits of advice have to do with creativity and being an artist:

• Always demand a deadline. A deadline weeds out the extraneous and the ordinary. It prevents you from trying to make it perfect, so you have to make it different. Different is better.
• Don’t be the smartest person in the room. Hangout with, and learn from, people smarter than yourself. Even better, find smart people who will disagree with you.
• To make something good, just do it. To make something great, just re-do it, re-do it, re-do it. The secret to making fine things is in remaking them.
• Art is in what you leave out.

And some of the more interesting ones are his disagreements with perceived wisdom:

• Following your bliss is a recipe for paralysis if you don’t know what you are passionate about. A better motto for most youth is “master something, anything”. Through mastery of one thing, you can drift towards extensions of that mastery that bring you more joy, and eventually discover where your bliss is.

One year later, Kelly has returned with 99 more bits of advice. I guess he couldn’t wait til his 99th birthday for it. Some favorites include:

• If something fails where you thought it would fail, that is not a failure.
• Being wise means having more questions than answers.
• I have never met a person I admired who did not read more books than I did.
• Every person you meet knows an amazing lot about something you know virtually nothing about. Your job is to discover what it is, and it won’t be obvious.

and finally:

• Don’t let your email inbox become your to-do list.

There is a small shift in Kelly’s 2021 list from his 2020 list, like a little more frustration with the world, a need for more order in the chaos. I wonder what his advice will be in a few more years?

via BoingBoing

Related Content:

Wired Co-Founder Kevin Kelly Gives 36 Lectures on Our Future World: Education, Movies, Robots, Autonomous Cars & More

The Best Magazine Articles Ever, Curated by Kevin Kelly

What Books Could Be Used to Rebuild Civilization?: Lists by Brian Eno, Stewart Brand, Kevin Kelly & Other Forward-Thinking Minds

Ted Mills is a freelance writer on the arts who currently hosts the Notes from the Shed podcast and is the producer of KCRW’s Curious Coast. You can also follow him on Twitter at @tedmills, and/or watch his films here.

167 Pieces of Life & Work Advice from Kevin Kelly, Founding Editor of Wired Magazine & The Whole Earth Review is a post from: Open Culture. Follow us on Facebook and Twitter, or get our Daily Email. And don't miss our big collections of Free Online Courses, Free Online Movies, Free eBooks, Free Audio Books, Free Foreign Language Lessons, and MOOCs.

Remote work expert David Tate wrote that when fearful CEOs talk about workplace culture, they’re really talking about workplace control. Their insecurities demand that the way work is done by employees is always visible, highly regulated and uses the methods executives prefer, rather than what’s best for everyone’s productivity. Remote work is seen as a threat to many CEOs simply because of their fear of change and resistance to progress. That fear leads to an irrational rejection of remote work, instead of a thoughtful examination of where it has succeeded and what can be learned.

In her May 6th Washington Post opinion article, I worry about the erosion of office culture with more remote work, CEO Cathy Merill makes two fundamental mistakes common among fearful executives. First, it shows an ignorance of alternatives, as many organizations have worked remotely for years before the pandemic and have solved problems she considers unsolvable. She may not prefer these approaches, but her lack of awareness of them is incompetence. Second, she is infantilizing her employees by presuming they are not capable of and motivated to be productive and collaborate even when the CEO can’t see them down the hallway. 

We are over a year into a pandemic and an era of great social unrest and uncertainty, yet Merill has chosen remote work, and not other likely psychological or cultural factors, as the singular reason why workplace performance has declined. And if this wasn’t enough of an oversight, her evidence against remote work consists mostly of examples from executive friends of their self-described management incompetence. 

She offered the story of an anonymous CEO with a new but struggling employee. Yet none of the leadership team did anything about it:

 A friend at a Fortune 500 company tells of a colleague who was hired just as the pandemic hit. He struggled. He wasn’t getting the job done. It was very hard for the leadership team to tell what the problem was. Was it because he was new? Was he not up to the work? What was the specific issue? Worse, no one wanted to give him feedback over Zoom when they hadn’t even met him. Professional development is hard to do remotely.

This is simply a management failure. Does this company not have telephones? Or email? Have they never worked with a vendor or client that wasn’t in the same building? They are responsible for helping this employee regardless of what technologies are available or not. This is inept management hiding behind technological fear.  

Merill estimates that 20% of work is helping a colleague or mentoring more junior people, extra work that she feels is impossible to do remotely. This is despite dozens of popular collaboration tools and mentoring programs that work entirely online. It also denies the dozens of remote corporations like Automattic and Citrix that have vibrant work cultures where these “extra” activities are successfully done remotely.

Merill and her peers might not like these alternatives, but she never explains why. She even goes so far as to suggest that remote workers should be paid less and lose their benefits, since in her estimation they will never be able to contribute in these extra ways. She effectively threatened her own staff through the article (she apologized later after her staff revolted).

If the employee is rarely around to participate in those extras, management has a strong incentive to change their status to “contractor.” Instead of receiving a set salary, contractors are paid only for the work they do, either hourly or by appropriate output metrics. That would also mean not having to pay for health care, a 401(k) match and our share of FICA and Medicare taxes

One quality of a great CEO is the ability to look into the future and show their organization the way forward. Instead of blaming employees, they take responsibility for solving problems. For every serious issue that arises they ask themselves what can I do or change in my own behavior that can lead my staff to a better place? They diversify their network to ask “who has solved the problem my organization is facing somewhere else and what can we learn?” Or perhaps most critical of all, they invite their own employees to participate in both defining the problem and exploring ways to solve it, instead of drawing lines in the sand and assuming the only way forward is the one that makes them the most comfortable.  

Technology is often seen as a silver bullet, oversold as the magic solution that can solve hard problems. This overestimates what a technology can do, as often it’s the management culture that is the real cause. But in the case of Merill, her CEO peers and remote work, technology is being used as a scapegoat. It’s the safe target to blame as it requires no introspection or accountability. Leaders that do this become fear-driven, allowing their competitors an advantage simply by exercising curiosity and seeking new knowledge. Smart CEOs chose to invest in their work culture and grow it for the future instead of hoping for the past to return. 

(Blogging was on hiatus because I've just checked the copy edits on Invisible Sun, which was rather a large job because it's 50% longer than previous books in the series.)

I don't often comment on developments in IT these days because I am old and rusty and haven't worked in the field, even as a pundit, for over 15 years: but something caught my attention this week and I'd like to share it.

This decade has seen an explosive series of breakthroughs in the field misleadingly known as Artificial Intelligence. Most of them centre on applications of neural networks, a subfield which stagnated at a theoretical level from roughly the late 1960s to mid 1990s, then regained credibility, and in the 2000s caught fire as cheap high performance GPUs put the processing power of a ten years previous supercomputer in every goddamn smartphone.

(I'm not exaggerating there: modern CPU/GPU performance is ridiculous. Every time you add an abstraction layer to a software stack you can expect a roughly one order of magnitude performance reduction, so intuition would suggest that a WebAssembly framework (based on top of JavaScript running inside a web browser hosted on top of a traditional big-ass operating system) wouldn't be terribly fast; but the other day I was reading about one such framework which, on a new Apple M1 Macbook Air (not even the higher performance Macbook Pro) could deliver 900GFlops, which would put it in the top 10 world supercomputers circa 1996-98. In a scripting language inside a web browser on a 2020 laptop.)

NNs, and in particular training Generative Adversarial Networks takes a ridiculous amount of computing power, but we've got it these days. And they deliver remarkable results at tasks such as image and speech recognition. So much so that we've come to take for granted the ability to talk to some of our smarter technological artefacts—and the price of gizmos with Siri or Alexa speech recognition/search baked in has dropped into two digits as of last year. Sure they need internet access and a server farm somewhere to do the real donkey work, but the effect is almost magically ... stupid.

If you've been keeping an eye on AI you'll know that the real magic is all in how the training data sets are curated, and the 1950s axiom "garbage in, garbage out" is still applicable. One effect: face recognition in cameras is notorious for its racist bias, with some cameras being unable to focus or correctly adjust exposure on darker-skinned people. Similarly, in the 90s, per legend, a DARPA initiative to develop automated image recognition for tanks that could distinguish between NATO and Warsaw Pact machines foundered when it became apparent that the NN was returning hits not on the basis of the vehicle type, but on whether there was snow and pine forests in the background (which were oddly more common in publicity photographs of Soviet tanks than in snaps of American or French or South Korean ones). Trees are an example of a spurious image that deceives an NN into recognizing something inappropriately. And they show the way towards deliberate adversarial attacks on recognizers—if you have access to a trained NN, you can often identify specific inputs that, when merged with the data stream the NN is searching, trigger false positives by adding just the right amount of noise to induce the NN to see whatever it's primed to detect. You can then apply the noise in the form of an adversarial patch, a real-world modification of the image data being scanned: dazzle face-paint to defeat face recognizers, strategically placed bits of tape on road signage, and so on.

As AI applications are increasingly deployed in public spaces we're now beginning to see the exciting possibilities inherent in the leakage of human stupidity into the environment we live in.

The first one I'd like to note is the attack on Tesla car's "autopilot" feature that was publicized in 2019. It turns out that Tesla's "autopilot" (actually just a really smart adaptive cruise control with lane tracking, obstacle detection, limited overtaking, and some integration with GPS/mapping: it's nowhere close to being a robot chauffeur, despite the marketing hype) relies heavily on multiple video cameras and real time image recognition to monitor its surrounding conditions, and by exploiting flaws in the image recognizer attackers were able to steer a Tesla into the oncoming lane. Or, more prosaically, you could in principle sticker your driveway or the street outside your house so that Tesla autopilots will think they're occupied by a truck, and will refuse to park in your spot.

But that's the least of it. It turns out that the new hotness in AI security is exploiting backdoors in neural networks. NNs are famously opaque (you can't just look at one and tell what it's going to do, unlike regular source code) and because training and generating NNs is labour- and compute-intensive it's quite commonplace to build recognizers that 'borrow' pre-trained networks for some purposes, e.g. text recognition, and merge them into new applications. And it turns out that you can purposely create a backdoored NN that, when merged with some unsuspecting customer's network, gives it some ... interesting ... characteristics. CLIP (Contrastive Language-Image Pre-training) is a popular NN research tool, a network trained from images and their captions taken from the internet. [CLIP] learns what's in an image from a description rather than a one-word label such as "cat" or "banana." It is trained by getting it to predict which caption from a random selection of 32,768 is the correct one for a given image. To work this out, CLIP learns to link a wide variety of objects with their names and the words that describe them.

CLIP can respond to concepts whether presented literally, symbolically, or visually, because its training set included conceptual metadata (textual labels). So it turns out if you show CLIP an image of a Granny Smith, it returns "apple" ... until you stick a label on the fruit that says "iPod", at which point as far as CLIP is concerned you can plug in your headphones.

And it doesn't stop there. The finance neuron, for example, responds to images of piggy banks, but also responds to the string "$$$". By forcing the finance neuron to fire, we can fool our model into classifying a dog as a piggy bank.

The point I'd like to make is that ready-trained NNs like GPT-3 or CLIP are often tailored as the basis of specific recognizer applications and then may end up deployed in public situations, much as shitty internet-of-things gizmos usually run on an elderly, unpatched ARM linux kernel with an old version of OpenSSH and busybox installed, and hard-wired root login credentials. This is the future of security holes in our internet-connected appliances: metaphorically, cameras that you can fool by slapping a sticker labelled "THIS IS NOT THE DROID YOU ARE LOOKING FOR" on the front of the droid the camera is in fact looking for.

And in five years' time they're going to be everywhere.

I've been saying for years that most people relate to computers and information technology as if they're magic, and to get the machine to accomplish a task they have to perform the specific ritual they've memorized with no understanding. It's an act of invocation, in other words. UI designers have helpfully added to the magic by, for example, adding stuff like bluetooth proximity pairing, so that two magical amulets may become mystically entangled and thereafter work together via the magical law of contagion. It's all distressingly bronze age, but we haven't come anywhere close to scraping the bottom of the barrel yet.

With speech interfaces and internet of things gadgets, we're moving closer to building ourselves a demon-haunted world. Lights switch on and off and adjust their colour spectrum when we walk into a room, where we can adjust the temperature by shouting at the ghost in the thermostat, the smart television (which tracks our eyeballs) learns which channels keep us engaged and so converges on the right stimulus to keep us tuned in through the advertising intervals, the fridge re-orders milk whenever the current carton hits its best-before date, the robot vacuum comes out at night, and as for the self-cleaning litter box ... we don't talk about the self-cleaning litterbox.

Well, now we have something to be extra worried about, namely the fact that we can lie to the machines—and so can thieves and sorcerors. Everything has a True Name, and the ghosts know them as such but don't understand the concept of lying (because they are a howling cognitive vacuum rather than actually conscious). Consequently it becomes possible to convince a ghost that the washing machine is not a washing machine but a hippopotamus. Or that the STOP sign at the end of the street is a 50km/h speed limit sign. The end result is people who live in a world full of haunted appliances like the mop and bucket out of the sorcerer's apprentice fairy tale, with the added twist that malefactors can lie to the furniture and cause it to hallucinating violently, or simply break. (Or call the police and tell them that an armed home invasion is in progress because some griefer uploaded a patch to your home security camera that identifies you as a wanted criminal and labels your phone as a gun.)

Finally, you might think you can avoid this shit by not allowing any internet-of-things compatible appliances—or the ghosts of Cortana and Siri—into your household. And that's fine, and it's going to stay fine right up until the moment you find yourself in this elevator ...

Time for a thought experiment! (For those of us who don't want to keep chewing on the sore that is the US presidential succession—if you do, please stick to this already-existing discussion: cross-contamination into this new discussion will be dealt with harshly.)

We all know by now that Elon Musk wants to appoint himself Pope-Emperor of Mars. As the world's richest man (currently, and only on paper—it's based on the Tesla share valuation, which is wildly inflated) and as the guy with the private space program that scooped 50% of the planetary civil launch market in the past decade, it's not entirely inconceivable. Evidently SpaceX hope to fly Starship to orbit in the next 1-2 years and land a Starship on Mars within this decade. Let's suppose it happens.

So then ...

Let's suppose that Musk's Mars colony plan is as viable as his other businesses: there are ups and downs and lots of ducking and weaving but he actually gets there in the end. All the "... and then a miracle happens ..." bits in the plan (don't mention closed-circuit life support! Don't mention legal frameworks!) actually come together, and by 2060 there is a human colony on Mars. Not just an Antarctic-style research base, but an actual city with a population on the order of 500,000 people, plus outlying mining, resource extraction, fuel synthesis, and photovoltaic power farms (not to mention indoor intensive agriculture to grow food).

Most of the city is tunnelled underground, using the rock overhead as radiation shielding. The radiation level to which citizens are exposed is nevertheless higher than in any comparable city on Earth: it's just the way Mars is. Workers in the outlying installations may be much closer to the surface than city-dwellers, and indeed most such plants are staffed on strict rotation by workers who are exposed to near-surface radiation levels for no more than three months in any consecutive Martian year.

Obvious aspects: cities are easier to heat and protect against radiation and provide with air and water, so housing is dense—think Singapore or Hong Kong density. High energy activities (eg. fuel and chemical synthesis, metal refining) and work with toxic substances are carried out sufficiently distant from the dense habitat that there's no risk of explosion damage. Musk's tunnel boring fetish turns out to be pretty useful when it comes to building a narrow-gauge mass transit system to move workers to/from these outlying sites, so there's a subway linking the city to most of its far-flung human-operated work sites.

More remote work is either fully automated, or largely automated but overseen by a small local dome full of canned apes with space suits. (Think the folks who go out to fix wiring faults on the big-ass solar farms, or to repair the robots that clean the PV panels after the dust storms pass).

The city is large enough that "self sufficient" is within sight. There's a teaching hospital and a university (a lot of educational material is distance-learning based but it's a hub that provides for tutorials/symposia and lab space). There's a semiconductor fab line, although the mine and refinery for on-planet sources of rare earth dopants keeps getting shoved back into the future: easier to import a few tons of gallium and lanthanum and so on from Earth every couple of years at a few million dollars a ton rather than spend 10% of your planet's GDP on a refinery that'll only break even after several decades. It's still a net population sink (most terrestrial cities were, until the industrial revolution: more people died than were born there) but there is a birth rate and it turns out that with a sufficiently good medical system babies can be born and raised on Mars without too many medical issues.

The population is overall young: nobody has lived there for more than 40 years so far, the oldest citizens are around 80 years old but somewhat more mobile than on Earth (having been born on Earth and now living in a lower gravity environment), and chronic illnesses that prove fatal on Earth over a period of years (eg. Alzheimer's) are either curable by this point, or result in short incapacitation followed by death. (In 2120 there will be a huge scandal and public commission of enquiry into the policy of "assisted dying" applied by the authorities to most of the first generation of colonists who lived significantly past their productive years, but that's another scenario. Let's just say, retirement of 12-18 months is tolerated: retirement over 24 months is almost unheard of because retirees are widely believed to "just give up".)

Most of the necessities of life can be manufactured or recycled with only minimal inputs. Pharmaceuticals, for example: modular chemical synthesis "bricks" can be plumbed together to produce different drugs flexibly. Lots of research aimed at disaster resilience on Earth—portable modular pharmaceutical factories, basically—turns out to be applicable on Mars. And mandatory pre-vaccination of colonists keeps the major human plagues from ever gaining a toe-hold in the new colony: there is and will be no mumps, flu, common cold, polio, smallpox, COVID19, HIV, or rabies.

The habitats are of course instrumented and surveilled exhaustively. Nobody wants to accidentally wander into a room that is anoxic because an air circulation fan packed in, or to sleep in a near-surface dormitory where a heater failed at night in winter. Nobody wants to asphyxiate in a cloud of sewer gas that burped from a waste tank with a blocked extractor pipe. Nobody wants to starve in a famine because the strain of fungi which play a vital role in some obscure phytonutrient recycling pathway got infected and crashed. And so on.

(Politics: Musk's autocratic dream didn't outlast his own lifetime and Mars is very locked down—ridiculously so, in the eyes of anyone accustomed to life on a planet with a self-sustaining biosphere. It turns out that dog-eat-dog capitalism is a bad fit for domed cities, which can't tolerate homelessness, civil unrest, or unmedicated schizophrenia. So socially it's a lot more like the Soviet bloc than the early 21st century EU or USA, albeit with much better planning/control/management and a governing ideology which boils down to lifeboat utilitarianism—"we're not building utopia, we're just trying to ensure survival for as many as possible in an intensely hostile environment (what were our grandparents thinking?)".)

So I'm going with the most optimistic take on a Mars colony in 2070 (short of invoking magical singularity woo and benevolent superintelligent AIs running everything).

What happens next ...

One of the regular biannual colony shuttles from Earth brings an unwelcome surprise: some of the essential supplies for the life support farms are contaminated with SARS-CoV-70, leading to an outbreak which starts among workers in one of the agricultural units (possibly a potato farm—h/t to "The Martian" here).

SARS-CoV-70 is the latest emergent vaccine-resistant mutant from the clade of respiratory coronaviruses descended from SARS-CoV-19. Failure to vaccinate to achieve global herd immunity in the 2020s resulted in these coronavirii becoming endemic, and with a large host population (natural immunity seldom lasts more than 1-2 years: often only months) it keeps throwing out mutant strains (eg. Lineae B.1.1.7, the more infectious strain of the original disease, which emerged in late 2020). When a new strain of SARS emerges which is resistant to existing vaccines, the World Health Organization coordinates another global emergency vaccine response, usually releasing a tweaked mRNA vaccine within 90 days: often only local lockdowns are necessary while the first doses are airlifted to the outbreak site. On Earth, SARS-type diseases are a recurring but well-understood problem: new outbreaks compare to the COVID19 pandemic of 2020-2024 much as a winter flu pandemic in the late 20th century compared to the 1918-22 Spanish Flu.

Mars is different.

Firstly, SARS-CoV-70 is vaccine-resistant: the Mars colony is a green field zone. Indeed, the policy of excluding diseases prior to emigration has resulted in a younger generation (20-30% of colonists) who are unvaccinated against anything, and probably didn't learn about historic plagues in history class (because why would they?).

Secondly, SARS-CoV-70 is comparable in mortality/morbidity and infectivity to the original COVID19: the one twist is that "long covid" post-viral damage is more prevalent, affecting up to 25% of survivors. The pattern is familiar, with 50% of long covid patients suffering serious organ damage and 30% being severely disabled after 6 months: "long haulers" on Earth follow a pattern familiar from other post-viral syndromes (eg. CFS) and may be impaired to the point of being unable to work for years or decades.

COVID70 is highly contagious, many carriers are asymptomatic, and it was spreading in close quarters for up to 11 days before anybody realized it had arrived.

However, there is some hope due to peculiarities of architecture on Mars. All rooms have pressure control, and where a hospital on Earth would have fire doors, a hospital on Mars has emergency airlock doors. Habitats in a Mars colony by default have to be able to lock down in case of a depressurization accident, and are compartmented like a submarine. Finally, per-room breathing gas control makes nursing support for patients much easier than on earth—you can crank the partial pressure of oxygen in the ICU right up (as long as you stay below roughly 28%, above which even waterlogged organic tissue is potentially inflammable).

An mRNA vaccine for SARS-CoV-70 is available off-the-shelf on Earth. Problem: Hohman transfer orbit windows are biennial, and the next one won't open for another 15 months or so. The hospital and university have limited flexible manufacturing capacity for mRNA vaccines—they're in constant demand, because they're highly effective against most cancers and cancer is a persistent health threat in the relatively high-radiation environment on Mars.

Figures: 60% of the population will suffer from cancer at some point in their lives, compared with 30-40% on Earth. Most cancers can be treated with a course of mRNA shots that teach the patient's learned immune system to recognize the cancer clone. 0.5M people ➙ 300K cases over 75 years ➙ the colony requires capacity to manufacture at least 4000 treatment courses/year, as a routine baseline for survival. So there is a vaccine factory and local expertise, and this can be scaled up, but to provide a COVID70 vaccine for the entire Martian population within a year would require at least two orders of magnitude more output.

So their choices are (a) wait 15 months for the vaccine shipment (and upgraded vaccine factory) to arrive from Earth, or (b) divert resources into lockdown, contact tracing, nursing, and jerry-building an emergency vaccine factory from equipment/expertise/parts on hand.

Oh, and note those 90 day crew rotations to the outlying fuel, refinery, and mining plants outside the city limits ...

Some additional parameters

(I will add to this section as other stuff occurs to me. Check back often!)

The UK in 2020 operates a very lean medical service that employs roughly 1.8 million people out of a population of 68 million. 1 in 40 is thus a low-ball estimate of the proportion of the Mars population who will be working in one or another area of medical practice—nursing, surgery, pharmacy, lab diagnostics, nutrition, general practitioners, physiotherapy. That means Mars General Hospital and satellite facilities employ about 12,000 people.

COVID family viruses kill roughly 1% of the total population, but health workers and school staff are disproportionately affected, with mortality/morbidity running at 300-500% of baseline.

COVID70 will, if unchecked, kill roughly 5000 martians ... but perhaps 500 of them will be doctors and nurses. Training new doctors/nurses is a 7-10 year process, and recruiting on Earth may be difficult in the wake of a pandemic hitting a closed environment.

The colony will be left with a legacy of maybe 50-60,000 disabled colonists, of whom perhaps half will recover in 3-9 months: the rest are extreme long-haulers. Repatriating the disabled back to Earth is not an option (many of the invalids would be harmed or killed simply by landing on a high-gravity world, after spending years or decades on Mars). You're going to have to work out a social policy for handling dependents. In contrast, the colony age distribution resembles contemporary China emerging from the one-child-per-family decades rather than any historic high birth rate/high death rate colonial model. Growing your own doctors and nurses—or care home workers—is a very long term project.

The question

You are the Mayor of Armstrong City, facing a variant SARS pandemic, and supplies and support are 15 months away. What do you do?

Alternatively: what are the unforeseen aspects of a SARS-type disease infiltrating such a colony?

And what are the long-term consequences—the aftermath—for architecture and administration of the Mars colony, assuming they're willing to learn and don't want it to happen again?

Discuss.