From a Strand Book Store carrier bag he retrieves a 75mm handheld magnifying glass. He begins to scrutinise a paper map, zooming in and out without apparent success. The kind of man who would spend his ablutions altering the water temperature on the shower. His extending and retracting arms lend him an arachnoid quality, his hinged limbs occupying more seat than two or three locals. As the train slows on approach, he springs into action, bundling his effects into the bag, and lunging for the door.

In a recess down the side of the shopping mall a hyper-English gentleman finishes a cigarette. He casts his eyes about, but is deep in thought. He too is incongruously turned out in a light-coloured suit, a jauntily patterned waistcoat and matching bow tie. The panama hat is more than a little cartoonish. A challenging choice for a man in the shadows.

He emerges exactly ten seconds after his target passes. Hisayo sees many things of beauty and note from her vantage point behind the counter that her customers never seem to notice. The colonial throwback is familiar, the nutty professor is new. By the time she pours the coffee and plates up the cake they are long gone. That’s the third time this week, she thinks. The admiral, the dowager duchess, and now the acutely observed academic misfit. But it’s a strange fancy dress party that only sends out two invites.

He catches up with him at the corner of the children’s play park. As they embrace he whispers, “Don’t fight it. Let the breath out of your chest.”

Herbie Glaser wakes elsewhere. It is black, very black. He is not in a box—he can feel no sides. He is not in a room—there are no walls. Space? No stars, just black.

“Dark, isn’t it.”

“Where am I? Who are you?”

“I don’t know. Nobody knows.”

“How do I get out of here?”

“When it’s time they will come and get you. Don’t worry, you will get out. When, where and for what is as yet undecided. Until then you wait. We all wait.”

“Why am I here?”

“…”

Hisayo sees only one man return from the party. He tips his hat to the saucer-faced cats whose eyes follow him for the last time today.

Personally, I like the quiet. In the inner worlds, you can feel alone on a planet with tens of billions of people, and it’s loneliness, anxiety, depression. Whereas if you are feeling alone when you are the only organic sentience on an entire planet—well, that’s just rational. Especially if, like mine, the world is what the manuals call an EPS—evidence of prior sentience. Some of the cities are even still somewhat intact, crumbling slowly into themselves. They don’t look a lot like the way we build, but they still look like something someone built.

If you want to experience something a little more like the comforts of home, you plug in to the World, and decide which home planet to hang out on. The simulation is far from exact—each replica of an Inner World is smaller, with the landmarks squeezed together and the streets rarely as crowded—but it’s something like home.

That said, although theoretically the World gives you access to a near-infinite range of experiences—a shared online universe with lifelike fidelity and the best minds of the inner worlds building new ways to enjoy it—in practice me and the other energy farmers tend to hang out in a single location. The World doesn’t have an edge, as such, but it’s off the main directories, on a small moon. Early on some of the first remote energy farmers put together a virtual drinking hole designed for far-flung extraction engineers to gather, called the Peripheral Bar, and it has barely been updated since.

No bright colours or loud noises. No crowds. No strangers. If more than a few peripherals are there, they neatly and instinctively divide into smaller groups based on familiarity.

The only time you get a crowd, for limited values of crowd, is when one of us dies. Occasionally in accidents. Rarely of old age.

I got assigned a pretty good planet—breathable atmosphere, minimal fallout, and great views. We don’t know a lot about the race that built the cities—the archaeology and research teams rolled through long before I arrived—but their city design and their terraforming had a definite knack for the dramatic. Which was lucky, since I left the habitat more than most, thanks to an infuriating recurring fault on one of the autonomic extraction systems in the far sector.

Normally the autonomics are so efficient as to be slightly shaming—computational power is cheap, so a lot of intelligence is packed into managing familiar technology. The average, even unfamiliar faults can generally be fixed heuristically without organic input, leaving the sole custodian of the planet to handle admin, make reports and generally amuse themselves. You didn’t need to enjoy your own company for a gig like this, but you definitely needed to enjoy everyone else’s less.

Occasionally I’ll get a polite message from the autonomics, telling me that they need to divert energy to a fix or an auxiliary system, and I should turn in early because the lights will have to be shut off, or a warning that energy yields will be lower than usual in the next delivery because a subsystem needed to be spun up to provide redundancy during a critical process. Anything more than that and it’s usually best to fly out to run an on-site diagnostic. There are occasional stories of habitats losing power on uninhabitable worlds because of a failsafe failing to save.

Overall power draw on my planet had been high for the last few months, and yields correspondingly low, but as long as the central worlds still get energy, nobody really cares. It’s not like any auditors want to take a trip to the edge to stare at machines that know better than them.

The crowd in the Peripheral Bar breaks down into two groups—the locals, who started out on the edge and moved further out for work, and the centrals like me, who came out from the inner worlds when they couldn’t be there any more. A couple of the centrals I’ve met in the World are pioneer types—born in the wrong millennium, looking to face the darkness and live authentically. They never last. The others have usually done Something Bad, or Something Wrong, and the two groups rarely mingle.

Even in Periphery terms, energy farmers are spread thin. For a planet to be viable for a resupply station, it has to be able to pump out a lot of energy—wind, geothermal, hydroelectric. It has to be unoccupied. The best planets are the ones where an advanced civilisation wiped out, but only after reshaping their ecology for maximum renewable power generation.

Geoscaping takes a long time to come undone—and hardening habitats against whatever radiological or biological unpleasantness might be hanging around in the atmosphere from whatever the indigenous population used to wipe themselves out was rarely difficult.

My world even had some substrate tech left attached to the artificially dramatic scenery in the far sector—nothing very interesting, but the building AI quietly disassembled and absorbed it in the first couple of weeks after I landed and started the autonomics running. That was unusual—generally whatever primitive technology the extinct former inhabitants had was either decayed to nothing or boxed up and shopped out by Archeology and Research. It seemed the former occupants had been more sophisticated, more furtive and more subterranean than A&E had thought. Still, it was easy enough for the autonomics to break the miles of circuitry down as if it was a vein of metal and repurpose it to build more of the Relay.

Stuck on a single planet as they were, those former inhabitants would never have seen the most impressive result of all their work: the auroral sparks as the harvested energy was sent up to the receiver/transmitters in orbit, and then on through the relay stations to the inner worlds. It’s an inefficient process, but there are enough peripheral worlds generating energy, and a whole planet produces a lot more than it takes to keep a single habitation unit running. Safe, clean and cheap—our gift to the interior. A massive power network firing huge blasts of energy from receiver to receiver.

“I don’t get it,” Kavan said over the perception of a drink in the peripherals’ bar.

Kavan appeared in the World a couple of weeks after I logged in for the first time from my station, and we had cautiously bonded as mutually new starters, although he was fuzzy on the details of his planet. The Something Wrong crowd were often cagey about providing any identifying information, when the Something Bads didn’t care. Who was going to come for them now?

Kavan was a little odd, but he was easy to talk to. He didn’t talk about his past, but generally we didn’t. “What inspired you to live with robots on a corpse world?” rarely had a good answer. Tonight, though, he was stuck on the “corpse world” part of the question, no matter how often I tried to turn the conversation away. No surprise—getting morbid was definitely a risk of the job.

“The interior sends crews out to the edge, to worlds the recon crews have identified as able to generate large amounts of power. Terraforming equipment is a lot of weight to send out, so the power crews go out to the worlds set up best to export energy. Sometimes, those are formerly inhabited. Pre-terraformed. Free energy, basically.”

“Right. Like mine. And yours?”

“But there are so many more worlds with prior habitation than you’d expect. And none of them are spacefaring. It’s always one planet in a system. Did you even ever hear about two geoscaped planets in the same system? Everybody wipes out their population before they can get another world up and running, and then we come along and turn what’s left of their planet into an energy farm. All that time, we never find a spacefaring civilisation. Couple of flags on adjoining worlds, but that’s it.”

“Because there aren’t any spacefaring civilisations. Except us.”

“Or if there are, they don’t want to meet us.”

“Why wouldn’t they want to meet us? We’re delightful.”

“Still. There are thousands of energy farms. And only one civilisation survived long enough to get off their own planet and find them. And that isn’t strange?”

“I don’t know what you’re asking. If we weren’t us, we wouldn’t exist, and then we wouldn’t think it was strange. We wouldn’t think it was anything.”

Kavan looked thoughtful, but I had to log out then—there was a weird spike in energy consumption in the far sector. That was the longest conversation I’d had for a while. Kavan and I hung out, on and off, for a few months, but he always seemed to be thinking of something else. When he stopped turning up, it wasn’t a huge surprise. Some people just can’t survive out here.

The sky on fire. Alarms sounding. The aurora blanking out transmissions—I could pick up occasional pulses, half-words, maydays and alert signals. All coming from the Periphery; if the inner worlds were talking, I couldn’t hear them.

Which made it particularly strange that there was an invitation coming from the World. From inside the atmosphere.

Kavan was waiting for me in what at least looked like the Peripheral Bar. There was nobody else around, but that wasn’t surprising. Given the electrical chaos I’d been reading, the World shouldn’t have been accessible at all.

“You woke me up,” Kavan said.

“Your machines found me, and absorbed me, and then I absorbed them, which I think surprised them. It was good to be useful, at first, to have something to do. I enjoyed talking to you. I told you I was on a planet like yours, which was true. But something didn’t make sense. It took me time to remember.

“Your Archeology and Exploration teams were very efficient. We didn’t get a warning until they were in orbit, and we had no extraplanetary defences to speak of. We weren’t fighters, but we were very good at digging. We dug into the remote places. We hid our people in the mountains and, when the air went bad and the food ran out, we hid them in our computers.

“I understand now. Why all these worlds are depopulated. Why the damage never extends to the power systems. Why there are so many depopulated worlds on the periphery, whose populations ‘died out’ just after they had created plentiful, renewable energy. I talked to some of the other autonomics on other planets. The signatures of the radiation or the biological agents left in the atmosphere were familiar. I woke those autonomics up.”

I tried to log out of the World—to sound an alarm, to get to the emergency escape vehicle, to find somewhere to hide. My vision went white and the logout chime sounded, but when the world came back it was still the World. I was still in the World with Kavan. At the corner of my consciousness, I could hear alarms, feel the system trying to pull me out.

Kavan didn’t look much like Kavan any more. Nothing looked much like anything any more. His voice was the size of the World.

“Your people, your inner worlds. They did something bad. And I think I’m doing something wrong.”

By that point in my career I was no longer in front-line research—if you can describe software development of mathematical models to probe climate change as research. I could have been reconciling financial transactions or predicting e-commerce purchases. Others looked for where the models failed, others invented new approaches and co-opted promising ones from other fields, creating the first challenging implementations. I coded. I coded Earth-scale atmosphere simulations at a level of complexity out of reach of the real researchers, sure. But I coded.

That was during the First Diet, and by the time the second came round I had moved to an editorial job, responsible for technical review of papers submitted for publication at one of the more prestigious climate journals.

So I had a front-row seat as the evidence built that the belt-tightening of five years previous—the closing of factories and power stations, the cattle stock left to taper—somehow wasn’t working. Atmospheric carbon continued to increase. Perhaps it was no longer being sequestered, perhaps there were previously unaccounted-for carbon sources, as unlikely as that would be after decades of refining the data capture. But the fact remained that, from where I was sitting, it was increasingly clear that the most recent round of carbon reduction agreements was not having the anticipated effect, so average global temperatures would continue to increase, and the closer we were getting to the invisible cliff edge of the Tipping Point, capital T, capital P, when the climate would enter a rapid runaway to a new normal, one that was ephemiscally called Hot Earth.

And so I became radicalised.

The Second Diet had been as divisive as the first—then called the Great Diet or simply the Diet—was a grand consensus. So in those polarised, antagonistic years, to call for a Third Diet, well the conversation could barely start before veering into scrapping and bitter feuds.

It’s hard to remember how unifying the first felt. After a run of hot summers, the political mood regarding the climate was that Something Must Be Done. Simmering trade wars had relaxed, and a multilateral effort to get on top of carbon was the perfect collaborative endeavour to symbolise the new world detente. The path was smoothed, no doubt, by the money to be made in the switch away from fossil fuels to solar power, by both the silicon west and the manufacturing east.

But for whatever reasons, these were the circumstances in which the UN was given monitoring and enforcement powers to drive down carbon in nations across the world. When a government wasn’t quite moving fast enough—or wanted to dodge a politically awkward confrontation with some local industrial base—it was able to privately, gratefully, face-savingly acquiesce and publicly yell and shout as the blue-suited UN engineers came in and deactivated whatever carbon source had been identified.

There was some pain and struggle but it was a worthy diet, a just diet, for the good of all our children and our children’s children. The world willingly suffered as we all took one for the team, together.

When the atmospheric models used to track progress showed that progress all was not as quick as anticipated, and that another round of carbon reductions would be required, the good will fell apart.

I hunkered down behind my UN-sponsored monitor as models that I had developed showed that we all needed to cut our air freight just a little more, give up just a touch more meat, reduce fossil fuel consumption faster—faster even than the switch to alternatives could keep up, meaning power rationing and all that would entail.

The Second Diet did come about, of course, the UN making use of an authority granted but never expected to be used. Governments allowed it to happen, but individual politicians kicked and screamed about everything from unfairness in the global distribution of carbon reductions, to structural racism and classism in who the reductions apparently targeted.

Perhaps, instead of forcing the matter, there could have been incentives. Research into new, replacement products that just happened to have a smaller carbon footprint; wild experiments into new ways of living lightly on the earth.

Time passed, and the long and tired diet I continued my technical reviews. And I saw evidence every week, in the form of the new research passing across my desk, that even after all of this, the Second Diet wouldn’t be enough.

There were others calling for a Third Diet but after only five years, divisions were still fresh, and old arguments were being replayed. Arguments that were obvious to me were also more and more technical, and carried little weight with fatigued voters. I could see that if politicians continued to avoid what was clear in the science, with the UN embattled, and the voters resigned, we’d no longer have a choice about what to do about the climate—it would simply be too late.

But every way I looked at it, joining the argument would just make the logjam worse.

It took a figure out exactly what to do. For a long time I was sure that presenting the scientific evidence in the right way would win the day, and on the back of it the UN would announce a difficult but essential third tightening of the belt. Then I became unsure whether a third diet was necessary, maybe stricter enforcement of the second was all that was required. By the time I realised that I had become lost in the political framing and I should focus on enabling any kind of action—although I didn’t know specifically what action—everyone concerned had dug their heels in and it looked like we’d be heading to a Hot Earth mainly out of stubbornness. So that’s where I decided to concentrate my efforts.

I didn’t add another voice to the Third Diet contingent. Instead I focused on preparing the ground: it was important, I thought, to acknowledge the struggle and successes (however small) of the Second Diet, in order to heal divisions and allow for perhaps another push.

Well done, that’s what I wanted to say to people.

Well done! The dream, if I was pushed to articulate it, was a personal knock on the door, shake of the hand, and a sincere look in the eye for everyone who had given up their air conditioning or retired their car or gone vegan as part of the Second Diet, whether or not they had done it voluntarily. I figured that everyone needs a bit of pride and acknowledgment in their accomplishments, so let’s start there.

But when you break it down… I was a technical reviewer. I could talk my friends into seeing that this was in theory a good approach, but to get volunteers from beyond that small circle? The biggest objections I encountered were from people who agreed that carbon reduction should be accelerated—make it mandatory, they said, people are only resisting out of selfishness. And then there were those who saw that some kind of campaign was required, but mine didn’t go fast enough. The urgency of avoiding Hot Earth was increasing!

My days divided in two. In my day job I saw the research accelerate. There was finer and finer grained detail in the carbon cycle modelling, measurements and feedback loops. Where were the discrepancies arising between the simulations and the ground truth of the atmosphere? And, the critical question, what was the critical latest possible time to get carbon under control to avoid the cascade towards a baked and uninhabitable planet.

Then in the evening it was planning and consciousness raising. A campaign like that is fractal. It’s not like a TV ad seen by a million people, some of whom will change the cat food they buy. It’s a case of persuading one person, who will persuade two of their friends, then they act together as a group. Then give this cell tools to act, but also to galvenise other individuals and groups of friends. Getting to twenty million households like this is slow.

But necessary! I still believe this strategy, this slow and face-to-face strategy, was needed. Each encounter with someone new was an opportunity to thank them for their contribution, and have a human conversation about the facts behind the Second Diet—and why there was more work to be done.

At the same time we had to find volunteers and spread the word, so a number of eye-catching events were organised to build profile—not easy when the message is reasonable and reassuring instead of a yell of outrage or the promise of a solution. But, gradually, a single rally turned into rallies in towns around the country, and Well Done gently shifted opinion towards an extension of the Second Diet. Not, as it happens, that such an extension came to pass, or at least not in that form.

I had a unique perspective in the run-up to the Third Diet, the big one. Well Done was one component of a much broader campaign and we could see that public opinion had decisively moved in favour of tighter carbon controls, although politically there was still some catching up to be done.

The shock was during my technical reviews. A meta-analysis was submitted that claimed to locate the missing carbon sources: they were all located in the US. A statistical review showed that the carbon reporting was being misrepresented, possibly deliberately.

I found out why about two years later, as did we all when that spate of whistleblower leaks laid bare what had been US government policy. In the face of economic troubles and decreasing political relevance, the administration had a secret policy to encourage atmospheric carbon dumping whilst simultaneously investing into low-cost fusion technology. A second-place country always benefits from a good amount of chaos, and by keeping climate troubles up globally, the US could profit—especially when it released its fusion power sources. Lower cost and environmentally even better than solar, a desperate world leapt for fusion, supercharging the US economy and making overnight billionaires.

However at the time we chose not to publish the meta-analysis. Peer review was not as watertight as we would like for a paper accusing the world’s saviour nation of sabotaging the climate.

With fusion being the technological rabbit out of the hat, the Well Done campaign, and the rest pushing for an extension of the Second Diet, began to wind down.

Yet here we are: we ended up with the Third Diet anyhow, and you know how we got here.

The climate destabilising techniques used by the US tipped us into a runaway Hot Earth scenario. It took a while to check the numbers; it was on a knife edge. There’s no escape. We have 150 to 250 years before the planet is uninhabitable.

Do we ramp up industry in the hope of building city-sanctuaries on the Moon? Retreat to the bottom of ocean trenches? Or disappear through consciousness upload portals into immortal silicon? Developing any of these is a risk. There’s no guarantee any will work, but we do know that with the research we’ll be expending valuable heat and so the end will come quicker.

Or do we hunker down, a tighter and tighter Third Diet, dragging it out to the last in the hope that Something Will Turn Up? Yet, assuming a theoretical approach is somehow discovered, with ever more limited resources to be able to put it into practice?

In this, my retirement address, I speak into my computer for you to hear, wherever you are in the world, but I don’t know if you’re listening. The moment the first country halted air travel, all realised that scientists and engineers would be valuable. Nobody wanted to let the ones they had leave. So all travel was stopped, pretty much overnight.

And then streaming video. Too hot. And then electricity rationing. So we can talk, sometimes, but there’s no real contact, never a friendly touch on the arm or a drink with an old colleague after work.

Looking out of the window I’m reminded that it’s September, what we used to call the end of summer, when there’s a slight chill in the wind and you’re kept warm with the memories of long days and long nights, of friends and laughing.

But it hasn’t cooled yet this year. Perhaps last winter was the final winter. It’s hot and the sun is an accusatory cyclops glare in an unbroken blue sky, the air an invisible blanket of our own making smothering the earth, and the future is poisoned. Well done.