"Take some of. Uh. My clothes. I guess."

How is he even without clothing. That's not a thing that is allowed to happen by the laws of nature, people who aren't very little infants being outside of their homes without any clothes.

Laws of motion only ever specify clothing as a specific thing when it comes to some Miracle powers, you know? 

She definitely doesn't! 

"Okay, thank you. But it might get torn or fall off when I'm going really fast."

Why's she nervous about lending him her clothes? Oh, right, probably gender taboos of some sort, whatever.

So, he doesn't, strictly speaking, have to carry her.

And close potentially transmissive contact would be unavoidable if he did.

But he's gonna have to constantly watch her to not get into contact with anyone for a week.

And if he prevented her being infected for 5 days, but infected her on the sixth, Huxley's Flu could again be stealthy for up to a week.

He's... cautiously optimistic on cooperating with her as a Miracle to self-quarantine. She seems ultimately reasonable and compassionate, not because but despite socially-instilled qualities that Transhumanization could erode.

There's a quiet, mathematical voice inside of him, that's saying that there's one simple and foolproof solution to the problem of not infecting anyone else through her.

It's gonna be ignored.

Okay. Let's stop being scatterbrained and think clearly.

Option 1: Try his damnedest not to infest her. Wait a week. Hope really hard for no stealthy slipups or outliers in the incubation term.

Option 2: Infest her deliberately. Wait out a week. She doesn't really have good chances, but if she could became a Miracle, it would probably be a massive positive in unimaginably many ways, considering the situation. But there's no ambiguity in the outcome. 

Option 3: Simply make sure she doesn't infest anyone, regardless of whether she is infected. No. Fuck that. Brain, why. She was being nice to me and she was helping me and is that how we want to contact a new world I don't think so.

...he could do Option 1 for a long time, like two months.

If he doesn't infest her in 5 weeks of consistent behavior, she's unlikely to be infested in the end of 6th.

And viruses don't randomly do incubation periods nearly that long, right? I don't remember hearing about it. But I didn't read any medical encyclopedias either.

Yeah, okay, that kinda solves every problem neatly. 

Inconvenient as hell, but what’cha gonna do? 

And, you know, there's still some risk of being wrong.

But Jean isn't gonna let tiny chances rule his decisionmaking when it, like, really fucking matters and the trap is obvious?

He'll need to actually calculate the necessary time, which he trusts himself to do. 

For now, though:

"On a second thought, I don't think carrying you is a good idea. What if I haven't infected you, and then accidentally do it while carrying?"

"...okay."

For your information, readers, the author is currently procrastinating on running calculations Malet is supposed to do off the top of his head in like half a minute. 

Jean remembers no good data for Huxley's flu transmission for cases comparable to his, and remembers little about the topic in general. 

The one good piece of data Jean definitely know is that under normal circumstances, the symptoms of Huxley's flu manifest in kids within a single week. If it didn't, it can be reasonably presumed safe.

Of course, this isn't ironclad. There are going to be exceptions... and Jean doesn't remember the range in which the exceptions can fall, and this actually matters when talking about patient 1.

Hence, a model splits into two submodels - typical cases, and tail cases.

We further split each of those into two - where Aliyeh gets infected before Jean properly treats the situation like a quarantine, and after that.

So, four cases.

Case 1, Aliyeh is infected from the start, and it's a typical case. The quarantine always catches this in a week.

Case 2. Aliyeh isn't infected from the start, and it's a tail case. We don't really know how long can the tails be, and we can hardly assume that a virus can as easily.sleep unnoticeably for ten years as it can for ten days -  medicine isn't really one of Jean's interests, but he have never heard of viruses behaving like that*, and he have never heard of kid asymptomatic carriers of Huxley's Flu (it's common in adults, though!). But neither he can dismiss the possibility totally. So to put a number on it, for a maximum tail length N, let's say that the chance of Huxley's flu having that length falls proportionately with N. And since that assumption seems somewhat optimistic, let's very pessimistically assume that the shape of each tail is flat, and if the maximum tail length is 20 days past the first 7, the chance of symptoms manifesting on the 20th day are equal to them manifesting on the first.

The resultant distribution forms a sequence which sum, rearranged, adds up to 1 + 1/4 + 1/4 + 1/9 + 1/9 + 1/9 + 1/16...  This then simplifies to 1 + 1/2 + 1/3 + 1/4 + ... which famously adds up to infinity.

So we're going round that infinity down to around four. And then normalize all that to 1. 

It's fine. It's not rigorous, but it's fine.

*even though they do, sometimes!

So, 27 days after the first week, the probability to be infected on that day falls down to about 1/1645 of getting it on the 1st day after the first week, and to about 1/3846 of getting it after the 7th but before the 27th day. And that's in itself unlikely, because usually a week is enough to eliminate reasonable doubt. 

Slightly more precisely, there's some probability A that Aliyeh was infected initially, and some probability B that her infection is a tail case, and if so, a quarantine of 7+27= 34 days reduces the impact of waiting an additional day by 1645 times from the 8th day, and an 8th day is already considered excessive. And according to the curve, the chances continue to rapidly, increasingly quickly drop with each day (even though they never reach 0 and even though they theoretically sum to infinity).

So if we double that whole 27 day period, the tail from the initial infection might as well not matter.

The third case is where Aliyeh is infected during the quarantine and falls sick within a week, as is typical. This is where things get kinda interesting.

Because Jean is pretty sure he can be really careful, but he also doesn't really know just how careful he is. Thankfully, "Aliyeh not exhibiting visible symptoms during the quarantine" is evidence that the quarantine is successful, and in this third case, it is also absolute evidence that it was successful up until week ago.

If Aliyeh is not exhibiting symptoms on day N, that means that on a day N-7, and N-8, and N-9, and so on up to 1, the quarantine was successful. Which means we can use straightforward calculation to estimate if it was successful on days N, N-1, ... N-6. Say we begin with a (really pessimistic) 50/50 smoothly uncertain prior for quarantine's success at each day, and use the rule of succession.

If so, after N days, N-6/N-5 is our certainty in a day of quarantine, and so (N-6/N-5)^7 is our certainty in an uninterrupted week of quarantine.

Then, to get our confidence in quarantine from 1/128 after 7 days to 19/20, we need... 142 days.

These numbers. Wow. Okay, that's not encouraging. But hey, 1/2 each day was just a totally ridiculous prior. Jean can give himself some credit and assume something more reasonable. Let's say, in a less pessimistic but still pessimistic prior, that initially whether or not Jean infects someone in a week of intentional quarantine is a coin toss, and so the initially daily probability of safety is 0.906.That's as if we start with... eight days of initial confirmation. Of 142 needed to go to 19/20.

So the lesson here is that successful quarantine brings depressingly little information about it's own further success... at least if the uncertainty was smooth. In actual case, there's weird uncertainty about uncertainty at play here, and there's a great relief if the whole quarantine plan works at all for 10 days, because that means nothing obvious is missing.

On the object level, Jean is pretty sure his systematic quarantine efforts wouldn't, in fact, be a total failure. What? He's a Miracle, he can manage.

And with that in mind, the fourth and most complex case, where it's both Jean's later slip up and tail distribution, can probably be ignored for now.

And that's all Jean is going to think up on this topic in a minute or so without pencil or paper or reference books or any experience in the subject.

This modeling has impressively low accuracy.

It's fine, it's not gonna matter anyway.

So, are we going upstream? 

I guess we are.

Aliyeh kinda lost the plot where Jean is supposed to be a demonic creature when he started to worry and then fainted and then told her more totally weird unrelated things that aren't blunt denial of danger. She's listening to Jean, who proposed to help her helpless escape and is kinda sorta acting as if he knows what he's doing, and vaguely signaling that you know what you are doing goes a long way in getting Aliyeh to follow.

The whole thing is gonna come to a head a bit later, of course.

While in Jean's head, epidemiology is kinda being superceded by questions like "Am I ever meeting my family again, by the way?"

Maybe I'm just insane and jumping to conclusions after meeting this one unlikely girl!