My name is L, and every birthday, I meet myselves. When I close my eyes each birthday night, I find myself walking down a hotel hallway. I reach my room, which I share with two others. Until a year ago, we three were the same person. From then on, our histories diverge.
On my 13th birthday, for example, my roommates and I were randomly chosen from all possible paths starting from my 12th birthday onward. We each had a year’s worth of new experiences to share with the other two.
Similarly, we also shared our first 11 years with the occupants of the other two rooms in the hall, having diverged from their two branches two years ago. There were three hallways on each floor, whose residents we shared our first 10 years with. Each hotel had three floors, each street had three hotels, and so on.
We call these meetings birthday parties. At the end of each one, I wake up back in the real world and forget about everything that happened, at least until the next birthday party.
It started off simple enough. On my first birthday, three infant Ls played with each other in a single room for a day, and on my second birthday, nine toddler Ls enjoyed themselves in a larger room for two days. It was on my fourth birthday, when 81 four-year-old Ls raised hell on earth in a hotel ballroom for four days, that I (we) had to get organized.
We found that the most people that could fit in a polite conversation was nine. So on the first day of my 13th birthday, I hung out with my one-year-group – three Ls who diverged on my 12th birthday. On the second day, I hung out with my two-year-group – nine Ls who diverged on my 11th birthday. On the third, my three-year-group was too big, so I hung out with eight other Ls, one from each one-year-group. On the fourth, I hung out with eight other Ls, one from each two-year-group, and so on. Each one of us could choose to leave the party and wake back up in the real world at any time, but most of us stayed for at least 13 days, and no one stayed longer than 30.
To summarize, on my nth birthday, there are 3n Ls present, (semi) randomly selected from the infinite set of all possible Ls starting from the initial conditions at the moment of my birth.
Well, at least there would be 3n present. The first one missing was on my fifth birthday. Though 243 Ls were expected, only 242 appeared. There was a car accident, we learned. There was an empty seat at the banquet hall. A branch trimmed, permanently. A hole in their one-year-group, three holes in their two-year group in a year, nine holes in their three-year-group in two years, and so on. All that they could have been would never be. We held a funeral in the hotel auditorium. From then on, it was tradition that each five-year-group hold a funeral for those lost in the past year.
Aside from these somber occasions, the birthday parties were typically laid-back and cheerful occasions. The best part is finding out what might have happened in another life (not what would have happened, since apparently the universe is not deterministic). When I was twelve, I remember my family considering moving to Texas for my dad’s job but ultimately deciding against it. On my thirteenth birthday, it was fun meeting an L who had moved to Texas, and each year afterward my branch of Ls would catch up with their branch. Eventually, we added extra days for events like the L chess tournament, Mario Kart day, the L art gala, other world movie night, and the L Olympics. Although some Ls got competitive at times, we were all friends in the end.
On my 16th birthday, however, enough Ls had started to care about politics to spark heated arguments, often spoiling the mood for the sizable majority who were still apathetic to such issues. We knew we were all the same person, so we generally tried to get along, and no one accused another of being evil or stupid. Still, we all wondered how the other side had become so misguided. It was social media, some of us claimed. No, it was our friends or teachers, others argued.
Despite the vast array of ideologies represented – communism, socialism, liberalism, conservatism, anarchism, nationalism, and even religious fundamentalism all had their adherents among the 42,387,708 Ls present – there was still a common thread among us. Nearly all Ls agreed that there was something deeply wrong with the world, and it was our duty to do something about it. Perhaps this was a something fundamental about being L, or at least a value locked in during early childhood. This too became a subject of discussion, rehashing the nature-nurture debate. Other Ls declared that it was because of our star sign or zodiac animal, or because God had chosen us to fulfill His mission on Earth.
Starting on my 18th birthday, we had the option to view statistics about us, including information that would have been impossible to obtain in real life. For example, it was determined that, at 18, the richest L was a millionaire who had won the lottery. A distant second was a finance prodigy who started with a few hundred dollars at nine years old. Although to be fair, the L who had started investing at nine now had 19,474 descendants, of whom only a few hundred were remotely successful, so maybe it was all luck.
In addition, it was found that Ls were 98.15% likely to live to 18, less than the typical rate of 98.87%. This was surprising, since Ls had no birth defects or genetic disorders, and they were typically healthy and risk-averse. There didn’t seem to be any reason we should be at greater risk of death than the average person.
We called these L statistics, and they were the coolest thing to Ls like me who were interested in science or math. While we had no way of influencing the real world and therefore were limited to observational studies, we had a supply of identical test subjects with identical backgrounds.
We quickly discovered limitations to the sample, however. In some ways, the sample was too homogeneous. All Ls were descended from one of the three Ls who diverged between birth and my first birthday. Similarly, all Ls were descended from one of the nine attendees of the second birthday party, and so on. This meant it was difficult to examine effects from early childhood.
By the 25th birthday, having such statistics also made it clear that some Ls were much more successful than others. A few had already become congresspeople, tech startup CEOs, or Nobel Prize laureates. On the other hand, some were homeless, some had alienated all their friends and family members, some were drug addicts, and many had committed suicide.
All of us dealt with this fact in different ways. Some were able to accept that there would always be people more successful than them. Others rejected the notion of success altogether and were satisfied with their lives. But for many, the problem was that the most successful Ls had had the exact same life circumstances as the less successful Ls, and yet they had been able to achieve much more with them. They had started in the same place and gotten farther; they had played the same game and gotten a higher score. Some believed this fact meant it was all luck anyway – if there was a game, it was chutes and ladders, and if anyone won (or lost), it was through no fault of their own.
Others believed that, unlike the differences between people in the real world, the differences between Ls could not be boiled down to genetics or parenting, since these were usually held constant. The most successful Ls represented the potential that all Ls had, and if you weren’t as successful, it meant you’d failed to live up to your potential because you weren’t as brave or diligent or smart.
In response, the former group claimed that if you weren’t as brave, diligent, or smart as some Ls, it was all because of chance anyway. Your parents read a different parenting book, or you had a different teacher, or your elementary school had tackle football. Chance also gave some Ls more opportunities or second chances. Your parents moved somewhere with worse public schools, or you didn’t meet the friend who would connect you with your first job, or your neighbors were partying the night before your GRE, or your university offered little support or understanding for your mental health issues.
Most people fell somewhere in the middle, deciding there were some things up to chance and some things up to your will – just like most people felt in real life.
---
I’m currently at the 30th birthday party, in the year 2030. We’ve expanded the list of events to include things like the L hackathon, the L beauty contest, the L Olympics, the L research conference, bring your family (pictures) to your birthday party day, the unluckiest L tournament, the L potluck, and its rival spin-off event, the vegan L potluck.
We’ve organized ourselves into clubs, including professional clubs like the L Bar Association and the L Chemical Society, hobby clubs like the L Football Club and the Bring Back WoW Server, political/cultural clubs like the L Conservatives and the L Buddhist Association, and identity clubs like the L Expat Union and the LGBT+ Ls. As Ls grow in number every year, these clubs typically develop sub-organizations under their main umbrella and develop ever more complicated org charts to keep track of everything.
I myself am an avid member of the L Actuarial Society, and when preparing a paper to share with them, I’ve stumbled upon a curious result. To determine what is responsible for the high death rate, I wrote some code in Viper to compare last year’s L death statistics with real world mortality, seeing if Ls are more or less likely to die of certain causes than the typical American. The first time I ran the code, its output said that the most common causes of deaths for Ls were:
LRank | ARank |Cause | LMort | Amort
1 | 1 | Unintentional Injury | 6.0050549 | 6.1490204
2 | N/A | Starvation | 5.0427960 | N/A
3 | 3 | Violence | 4.8984499 | 1.3946951
4 | 2 | Suicide | 2.3719091 | 2.0288265
5 | 14 | Other Inf. Disease | 1.3840906 | 0.0615072(Note: L refers to L statistics, and A refers to American statistics, averaged across worlds. Mortality is given in deaths per 100,000)
Unintentional injury and suicide didn’t seem out of the ordinary, but the other causes were surprising. So few Americans die of starvation that it’s not even recorded anymore, hence the N/A. And except for the few Ls that became soldiers or police officers, and the weird L that ran off to join ISIS, few (surviving) Ls have ever experienced violence.
And what could “Other Infectious Diseases” mean? After digging deeper, I found that the category was dominated by either novel diseases not recorded anywhere else or novel strains of known viruses. This even included extinct diseases, like smallpox. What could be causing this? I know some Ls that travel to remote locations as development economists, doctors, missionaries, or ecologists, where they might be exposed to new diseases. Other acquaintances work with animals as biologists, animal rights activists, or veterinarians, which might expose them to zoonotic diseases. But a quick search shows that these professions are nowhere near sufficient to explain the large fraction of exotic disease deaths, and neither are the natural pandemics that have become a fact of the 21st century.
Now beyond curious, I leaned forward in my chair and began looking into the starvation deaths. Most Ls never missed meals, but their diets would start becoming more and more irregular a few weeks before their deaths. They would start missing meals for days at a time, and when they did eat, it would be a haphazard mix of nonperishable food. Canned meat would even start showing up in the diets of vegan Ls. Desperate, I looked at their financial situations for an explanation. Eventually, a pattern emerged: their earnings would abruptly drop to zero, their savings would either be all withdrawn immediately or remain untouched until their deaths, then they would miss meals, then they would stop eating altogether, and then they would die.
I altered the code to show how these causes of death changed over time. I held my breath as it compiled and ran. I saw the graphs render in real time, the lines being drawn from left to right. Starvation was noisy, but it consistently hovered around 0.04% per year from 2000 to 2030. Diseases started near zero in 2000 and stayed that way until the late 2010s, when the red line lifted off the x-axis like a plane off a runway. My heart pounded as the violence graph rendered. The blue line started under 0.05% per year, zipping past the markers for 2005, 2010, 2015, 2020, and 2025 without changing at all. I was wondering if I had made an error with the code when, in the mid 2020s, the blue line started climbing upwards, rising faster and faster.
A shiver ran down my spine as I realized the implications. Something is killing Ls.
I quickly went back to the violent deaths database to see what was going on, writing code to sort by cause of death per year. Until 2028, violent deaths were dominated by nuclear strikes, with homicide a distant second. But starting in the mid 2020s, the nanoweapon subcategory began increasing from near zero, overtaking fire/burn, then cut/pierce, then firearms. In 2029, they had become the largest subcategory of violent deaths.
Bypassing the process of writing and submitting a paper, I sent my findings directly to the president of the L Actuarial Society, an accomplished L who had become one of the board members of the real world Society of Actuaries. I received a reply within the minute.
“Dear L ACACABABCBACABBBACCCCCABAACCBC:
Thank you for bringing this to my attention, and kudos for noticing it. Many other Ls have noticed this alarming trend, and not just the actuaries. The birthday party planning council is planning to hold an emergency conference on this matter tomorrow at 8:00am at the convention center at this location, which you are invited to view as a guest. If you’re interested in attending as part of the Actuarial Society delegation, please apply here. Otherwise, feel free to join us in conference room 23321 to discuss this with the other concerned actuaries in the meantime.
Best,
L ACACABABCBACABBBACCCBBAABCBAAC
This was an automated email triggered by and generated from an analysis of your email contents, such as the keywords ‘emergency,’ ‘nanoweapon,’ and ‘violent death.’ The model estimates that this email is 94.31% likely to be a suitable response to your query. If it was not, please contact the L Actuarial Society helpdesk to get in touch with a human representative.”
I was surprised by neither the automated response nor the fact that enough actuaries had already discovered this pattern to warrant the president tasking their email system to provide such an automated response. There were tens of thousands of Ls in the Actuarial Society, after all, and we all thought alike. And given that there were 198 trillion of us, we automated what we could. I RSVPed for the emergency conference and made my way to the conference room.
---
There were over 80 billion attendees at the emergency meeting, so we used the “extremely large delegated conference” format. I’m omitting some details, but generally the attendees sat at semicircular tables of 9, and there were 27 tables in each theater facing the stage. The current speaker would appear on every theater’s stage at once, so all the attendees felt like they were attending a real conference in person.
I had spent most of the night updating my code and formatting my results, as had most of my actuary colleagues at the table. We excitedly shared our findings despite the exhaustion. After several minutes, an L emerged from backstage and walked up to the microphone, and our discussion quickly wound down. A powerpoint lit up on the screen behind them.
“Hi everyone. Thanks for coming. I’m here on behalf of the L Commission on Catastrophic and Existential Risks. If you’re here, it means you’ve noticed that an alarming number of Ls have been dying in violent nanotechnology attacks – congratulations. These trends were first discovered by my friends at the L Statistical Association and the L Actuarial Society, who independently found that Ls were dying of surprising rates to starvation, violence, or diseases – specifically, unknown diseases or diseases thought to be extinct, like smallpox. In response to this, the planning council set up the Commission to look further into the matter.
“Before I go any further, let me explain what catastrophic and existential risks are. Global catastrophic risks are things that threaten to derail civilization or lead to its collapse, affecting our long-term future.” A list of risks appeared on the screen. “Existential risks are more severe – they’re defined as things that threaten to make humans extinct, or at least permanently limit their potential. Threats like unaligned AI and engineered pandemics. In every world, people like me have been sounding the alarm about existential risks.
“In my world as early as 2020, the philosopher Toby Ord estimated that there was a one in six chance that humanity would become extinct by 2120. Thanks to the research done by the Commission, we have found that the true rate of extinction, from 2020 to 2030, has been 1.02% so far. This naively works out to 9.79% by 2120 if you assume the risk stays constant, which it almost surely won’t. This also doesn’t count worlds in which civilization has collapsed but humanity hasn’t become extinct yet, by the way. I’ve spent my entire career working on preventing these risks, as has most of my ten-year-group.
“But how do catastrophic and existential risks relate to these causes of death? We found that most deaths from starvation and some from violence were due to the breakdown of society, whether by nuclear war, economic catastrophe, or other factors. Nuclear war was also directly responsible for a fair share of violent deaths, given that Ls are likely to live in large cities. These are non-existential catastrophic risks, although they can indirectly increase the risk of extinction.
“Most deaths from violence were from nanotechnology – the result of unaligned AI or, more rarely, nanotechnology accidents or misuse, which both typically result in extinction. And the novel diseases were typically engineered pandemics, although a few were naturally occurring. These sometimes resulted in extinction.”
Of course, that explained why Ls seemed more likely than the average American to die of nanoweapons or exotic diseases! Our notion of “average person” comes from the average of data from each birthday party attendee’s world. If an L dies, their world’s mortality data isn’t included.
If the fact that an L dies typically has little relation to how likely people are to die in general, this isn’t a problem. But if some catastrophic event kills every human, it affects both. We only find out that the L doesn’t make it to the birthday party, but we don’t find out that everyone else also died, and that doesn’t get included in the actuarial tables. We don’t get data from worlds where everyone is extinct.
And in fact, anything that meant people were more likely to die in general also meant L would be more likely to die, which meant it was less likely that this world’s lower average lifespan would be included in the data. Classic survivorship bias.
“If you think about it, these catastrophic risks have been a fact of life for a long time. Remember L AAAABBCC, who was seven when they became the first one of us to experience and die from a nuclear war? Or L ACACBBBCC, who was nine when they became the first to die in a pandemic? And at every party since then, almost everyone has known someone in their five-year group or below who’s died from one of these risks. Some of us are even survivors of global catastrophes, like those of the Apocalypse Survivor’s Association. A few of us witnessed a global catastrophe unfolding just before coming to this birthday party, and they could very likely die as soon as they leave.
“But while we do have some firsthand accounts of survivors of non-existential catastrophes and current victims of slow-moving extinction events, nanotech catastrophes are different. In order for an L to make it to the birthday party, they need to have witnessed the catastrophe, then gone to sleep while it was in progress. But nanotechnology typically kills everyone very quickly, and none of the victims have been able to make it to the parties so far.
“I’m sure you all want to know how we can prevent such risks. But as you are well aware, there’s nothing we can do about it from the birthday parties. These findings will not be communicable with the real world.”
Indeed, we had learned long ago that it was impossible to influence the real world from the birthday parties, since we never remembered the parties when we woke up. And believe me, we tried everything. Not only were memories lost, but so were tattoos, scars, bruises, Pavlovian responses, or muscle memory. Even the internet was somehow read-only, which baffled all the computer scientist Ls.
“If you could remember these findings in the real world, I would recommend looking into supporting existential risk prevention efforts when you get home, either as an advocate or a donor. There are lots of organizations working on AI safety, biosecurity, nuclear issues, and other catastrophic risks that could always use more expertise and funding. That’s what I’ve been doing in my own world. I would plug my own organization, but it probably goes by different names in different worlds.
“Instead, all you can do is hope that they succeed by themselves in your world. What you can take away from this is that there is a 0.2% chance that you won’t be back next year, instead of the 0.09% ‘typical’ for 30-year-olds. This may be higher or lower for you depending on several factors, like how advanced biotechnology or AI is in your world, what the geopolitical situation looks like, and so on. Let your two-year groups and your friends know, and you can decide what you want to do with that information.
“Right after this meeting, the Commission is holding a Q&A to answer your submitted questions, if you’d like to learn more about their methodology or about what a global catastrophe would look like. The Post-Apocalyptic L Association and the Apocalypse Survivor’s Association will also be holding their own open meetings to share how their worlds ended, how to know if your world is close to an apocalypse, and what to expect. Although, keep in mind they are only describing worlds with survivors. Lastly, the L Mental Health Union and the L Philosophical Society will each have their own workshops to help you process this information. The schedules for these events, as well as many others, can be accessed from the invitation for this event.
“Good luck, everyone,” said the speaker. They did the L salute – extending the right thumb sideways and index finger straight up, then raising the back of the hand to the forehead. “I imagine the Eternal Birthday Club will see a rise in membership this year.”
A few chuckled at the dark joke. The Eternal Birthday Club was made up of Ls who chose to stay at the birthday party for a long time. They were typically Ls with terminal disease who knew they were unlikely to survive another year and who wanted to live as long as they wanted at the party until they came to terms with their death and left to meet their fate. One L had spent 6000 years there, although we never learned why because they had no descendants the next year.
If there was a 1 in 500 chance that some catastrophe would kill me in the next year, maybe I should stay a bit longer. I was still considering this when the lights slowly came back on.
“So,” said one of my colleagues, breaking the silence at the table, “who saw that coming?”
(For convenience, I will refer to my colleagues by their position on the table relative to me, clockwise ascending. The L who just spoke is L5.)
“I’m a little embarrassed I didn’t put two and two together,” I admitted. “I didn’t know AI and pandemics were so dangerous, but I should have recognized that survivorship bias would be an issue given, you know, stuff like nuclear war that affects everyone.”
“Yeah, that’s true. I’m sure there’s a relevant statistics paper somewhere where they account for this to find the apocalypse-adjusted mortality, or something,” said L1. “But I’m still not so convinced that it requires any change to what we’re already doing. An extra, what was it, 0.1% chance of dying every year? It doesn’t seem so significant. At worst it’s the same rate as if we were 40 years old. I mean, as actuaries we especially are aware of the fact that we have a chance of dying every year. And yet, we live life. This doesn’t change anything.”
“But it’s only going to get worse, isn’t it?” said L4. “In the worlds where AI or nanotech or biotech hasn’t gone bad, they’re still getting more advanced and influential every day.”
“Not only that,” L7 added, “everything else is getting worse too. If it isn’t that, it’ll be the US and China at each other’s throats, or climate change, or populism, or something.”
“Ok, sure. Say the accumulated risk makes our expected lifespan 10 years lower than it says on the actuarial table,” said L1. “But be honest, before attending this presentation, what did you think the yearly risk of civilization collapsing was? I personally thought there was a 0.5-1% chance of a nuclear war every year, so to me these findings are a relief. Were your estimates about the same?”
Five of us had the same or higher estimates, but three of us had thought it was less than 0.1%.
L1 continued, “Yeah, see? We already knew there was a chance that the world could end, and this affects every decision we make, like whether to save for retirement, have children, or join the Eternal Birthday Party. So even though we may be more likely to die than the actuarial table says, we are actually less likely to die than most of us already believed.”
There was a pause for a minute. Then L8 said, “Math aside, and personal risk aside, is this really something worth working on, if you could remember in the real world?” Upon seeing the others’ confused faces, they explained, “The speaker said this is really important to work on, and if we could remember we should help prevent pandemics, or AI, or nuclear war. But what can you really do to prevent these kinds of things? I think they are so intractable that, if your goal is to do the most good, you should do something else. For example, donate $10,000 to prevent vitamin A deficiency and save a life with high certainty.”
There was another awkward silence. Many of us at the table had essentially sold out and given up dreams of changing the world in the last 10 years. We had become actuaries, after all, not activists or leaders or scientists. We had chosen to use our math skills to pursue a career that was stable, high-paying, and perhaps even interesting and essential, but not world-changing. And it wasn’t entirely our fault – sometimes life, kids, or sick parents changed our priorities. Some of us still tried to stick a balance by donating a significant fraction of our income. I personally still found it awkward when meeting Ls who spent their lives fighting for what they believed in, kind of like how I imagined it would turn out if my grandparents traveled back in time 70 years and met themselves as hippies.
“I think,” piped up L2, “that the stakes are high enough where it still makes sense to work on preventing catastrophes. It’s a situation where the odds are low but the payoff is high. Yes, a single person might barely make a difference, but if they reduce the chance of human extinction by just a little bit, in expectation they could be saving many lives. Yes, I know it’s a Pascal’s wager thing. But depending on the number of lives at stake, and the amount you could actually reduce the risk by, it could plausibly turn out that when you do the math, preventing catastrophes is either way more efficient or way less efficient than conventional ways of changing the world.”
“How would we find out these numbers?” I asked. “That could be true, but until we know we’re all just speculating.”
“We could look into the L statistics,” said L8. “See whether the worlds in which an L works on preventing existential risks truly have a lower extinction rate than the baseline.”
“I think the difference they make is so small that any difference will be swamped by noise,” suggested L2. “But we should try it. Is the data published yet?”
“There might be a confounder,” said L4. “If a world has larger and more successful risk-preventing organizations, then L is more likely to be part of those communities. Those worlds could be less likely to face extinction, but it won’t be because of L alone.”
“Then we could just shift our attention towards studying the effect of having large risk-preventing organizations, and see whether they prevent extinctions.”
“How would you measure that? Even if you could, that itself could have problems. Maybe these kinds of organizations arise in response to a world having advanced AI, or pandemic near-misses. You’d need to find something that causes these organizations to become large and successful, yet has no other correlation with the risks themselves.”
“Sorry, I was just thinking aloud. You’re right, you would have to– wait, when do the next events start? Ooh, the Q&A’s about to start. Do you guys want to go together?”
---
I ended up staying at the birthday party for an extra three months, learning about the risks and pondering whether it was worth working on them. On the day I decided to leave, I took a seat under a tree atop a large hill, modeled after the one at my college campus.
At previous birthday parties, I had wished to remember many things. When I was 21, my two-year group convinced me that my relationship was toxic, and I hoped to remember their advice when I woke up. I didn’t, but it ended up working out. When I was 24, I was so moved by my five-year group’s funeral that I resolved to call my grandparents the first thing after I returned. I didn’t, but I ended up near their house by chance on a business trip that year and paid them a visit. When I was 27, Lcolholics Anonymous made me realize all the damage alcohol had done to me and those around me, and I hoped to retain my newfound sobriety in the real world. I didn’t, but somehow a friend introduced me to a similar group. Maybe there were some things we could remember deep down, or maybe there was a higher being after all, or maybe it was all chance.
Either way, it meant that this tree was very popular among my nine-year group and below, and it was difficult to reserve a spot there. This time I was one of the last to leave, so I had it all to myself.
I watched the familiar skyline, which I now only saw in memories and birthday parties. The sun rose from the mountains in the distance, piercing through the smog. I took a deep breath and closed my eyes.
I wish to remember…
Hope you enjoyed reading! Any feedback would be appreciated. I will literally Venmo you $5 for it (PM me).
Also, I submitted this the night of the deadline, but it wasn't moderator-approved until the morning after. Sorry for procrastinating
Warning: I'm an unrepentant perfectionist, incredibly nitpicky, and way more suited for describing flaws in things than for celebrating the good parts. I think I'm getting better at the latter, but, uh, don't be too discouraged.
To sum up: very interesting and well-explored concept, but the narrative focus is misaimed, and the management of readers' expectations needs some work.
No payment necessary: your willingness to pay serves as a strong enough signal of your interest to improve for me to take the time to respond.
Thanks for the detailed feedback! I found it very helpful and not nitpicky at all.
I struggled with finding a balance between explaining everything and letting the readers figure things out. My biggest worry was that it would be incomprehensible without the long exposition, but based on your response I should have run this by more test readers.
Regarding the ending, I definitely felt the speech and aftermath was the weakest part of the story, and I like how your suggestions would add tension and stakes like you said. When I have time, I'll see how I can make those work and edit them in.