I’m no longer on Quora, but I used to be. Here are some of my favourite answers.
Author
Chris von Csefalvay
Published
13 April 2022
I used to be on Quora. I’m not anymore. But I used to be. And I wrote some answers. Here are some of my favourites, categorised by field… sort of.
AI
If I wrote an AI and trained it to score 200 on a standardized IQ test, what could a human of 100 IQ still do better?
There’s something called Moravec’s paradox, which I’d sum up as follows: stuff that’s difficult for humans is not the same as what’s difficult for a computer. It is relatively trivial to teach a computer to be relatively good at chess or checkers or even go. On the other hand, getting a computer to master, say, bipedal motion is actually very, very tricky. There are multiple reasons behind this, including that the human brain is the pinnacle of an evolutionary pathway that spans the course of millions of years of optimisation for certain kinds of tasks, such as bipedal motion. We are an ‘opinionated’ system: we evolved with the idea that we’ll be solving certain kinds of problems, and not others. The evolutionary incentive for chess has been relatively modest compared to the evolutionary incentive to be able to run away from a sabre tooth cat or some prehistoric predator. Our brains have evolved through millions of years of exposure to certain senses: vision, balance, proprioception, pressure/pain, and so on. Without so much as wasting a single thought, our brain performs the process of sensor fusion and generating an operating model of where we are in space with accuracy that is really not trivial for computers even today. And we’re really good at this before we’re old enough to know just what an amazing nervous system we have. You really do not realise what a fantastic thing the human central nervous system is until it starts malfunctioning.
Will true AI come before the ability to completely simulate a human brain? Is there a difference?
Simulating the human brain is a fairly pointless exercise. You see, the human brain is a functionally extremely diverse thing. You see this best when you consider what damage to it can cause: some brain injuries cause mood disorders, some cause memory disorders, some cause aphasia (speech disorders), some cause issues with mobility and motor control, some cause severe disorders of consciousness like persistent vegetative states and minimally conscious states. The brain is basically a bunch of histologically largely similar tissue that does a lot of different things. There’s not a lot of point in simulating these functions.
A bigger problem, though, is that some things are a lot harder to ‘simulate’ than others. This is something widely known as Moravec’s paradox: it’s a lot easier to teach a computer to play chess at a pretty high level than to teach it simple bipedal human motion. For this reason, simulating the human brain has generally been seen as not all that much of a priority. Connectionism is great, but it’s largely an inspiration, not a one-to-one reality. Many get this wrong because of concepts like neural networks and artificial neurons. In reality, a neural network is an abstraction to help humans. On the code level, there aren’t individual neurons or layers, there is just a bunch of relatively trivial linear algebra going on. I know I’m oversimplifying a little, but that’s the general idea.
Why is it important to be concerned with the health of other nations when it comes to global health issues?
Here’s an illustration. I live in Denver, a city with a major but not first-in-line international airport (e.g. not like Heathrow or LAX, but a pretty nice airport no less, despite the creepy murals). Just as a thought experiment, I sometimes pick random places on the planet and look at how long it would take me to get there. It’s virtually never more than 48 hours.
To prove this point: it’s a little before 8am on a Sunday here. I can be:
…in Ürümqi, China, in about 35 hours, …in Makassar City, Sulawesi, Indonesia, in about 33 to 42 hours, …in Maputo, Mozambique, in about 29 to 35 hours, depending on airline, …in Itahanga, Mato Grosso, Brazil, in about 44 hours, if I drove straight from Manaus, or 34 hours if I flew into Brasilia.
I do these thought experiments all the time when I wonder about the thing you’re asking yourself. Because if I can get there in 48 hours, a pathogen from thereabouts can get here in the same time.
Do you believe that there could be more cases like the one found in Ohio, where someone has been infected with COVID for an extended period of time?
Absolutely. In fact, I’d be extremely surprised if that weren’t the case.
Let me take a slight detour here to a concept that might be more familiar. You’ve heard of supershedding and superspreading, right? We’re instinctively geared to consider things like that to be extraordinary phenomena, the same way some people who could hold an HIV infection without developing AIDS were known as “elite controllers” for a long time (today, the term is “long-term non-progressor”, which is much less cool, sadly). But the reality is, they’re not. They’re the (often quite fat) tails of a distribution. Here’s something that doesn’t come to you naturally, something you need to force your brain to wrap itself around as you start doing epidemiology: all the numbers you encounter talk about the central tendency, but say relatively little about the distribution.
Let me expand on that a little. What’s the mean infectious period of COVID-19? If you ask someone who has a relatively good grasp of the subject, they’ll say it’s around 9–10 days, with an interquartile range of around 4–12 days, give or take one day each way. That term – “interquartile range” – means that given a sample, half of the values will be in that range. Or, more importantly, half of the values will be outside that range.
Native Americans were devastated by European diseases brought by colonizers, but were the colonialists impacted by new diseases not found in Europe?
Yes, with a significant difference: the Americas had a population density nowhere near that of urban Europe, where many of the colonists came from. Europe is tiny, and very interconnected. The result of that is that pathogenic diversity is going to be off the charts. It’s not even that Europeans brought various epidemics into the Americas, whose populations were largely naive to those pathogens (a phenomenon widely known as a ‘virgin soil epidemic’, much as I loathe the term) – it’s that they bought hundreds of different versions of it. They brought that into societies that, at least where North America was concerned, lived in a way that is very significantly more resilient to pathogens: in small(ish) groups, often as part of extended families (i.e. often similar profiles of immunity), spread out. In a profound sense, Old World epidemics are the result of a peculiarly (but not exclusively) Old World phenomenon: the Big City (“big”, here, being generally relative to total populations). When colonists moved to the Americas, they bought a piece of that Big City’s pathogenic diversity with them, to populations that have neither been exposed to that particular microbiome, nor have generally been under the evolutionary pressure to select for better genetic predispositions to develop immunity to a wide range of pathogens, since that just wasn’t much of a thing in the New World.
The flu has been around since 1918. Why are the vaccines against Covid, which is much newer and more harmful, so much more effective than the flu vaccines? Is there any hope for a ‘universal flu vaccine’ (as there is for a ‘universal Covid vaccine’)?
Influenzaviruses have been around for much longer than since 1918. Greek physicians from the 5th century BC have documented what is almost definitely influenza. Influenza A and B viruses have differentiated around 2,000 years ago, but both trace their origins to a common ancestor around 8,000 years or so ago, give or take a few thousand years (molecular clocking gets tricky after some time). More importantly, however, influenza has a level of diversity that far exceeds SARS-CoV-2.
Influenzaviruses are actually not a single pathogen, but four genera of the family orthomyxoviridae, each of which comprises a single species: we refer to these as influenzaviruses A, B, C and D. These are viral species in their own right. A, B and C are human pathogenic, and A is responsible for pandemic influenza. Influenzaviruses then break down into serotypes. When you see something like H1N1, that’s a subtype – it denotes the type of two major surface proteins, haemagglutinin (the number after H) and neuraminidase (the number after N) expressed by the virion. Influenza A virus is incredibly genetically diverse, and mutates rapidly. This creates a kind of diversity and versatility that makes it hard to vaccinate against.
Why is “patient zero” called “patient zero” and not “patient one” to denote the first person with a disease?
Actually, it’s not. It’s a popular science term, nobody in epidemiology uses it. The correct term is ‘index case’ or ‘index patient’.
“Patient zero” comes from a contact network study by Auerbach et al. (1984) on HIV:
Auerbach’s contact network
I have a fervent dislike for the ‘patient zero’ term, both in colloquial usage and in epidemiology (where it’s thankfully quite rare these days), so make sure you read this answer in full. Incidentally, we’re looking at this contact network in Computational Note 2.10 (page 75) of my book.
Why are infectious diseases with high mortality rates not as infectious as those with lower mortality rates?
There’s a theory that underlies this observation, widely known as the ‘avirulence hypothesis’ aka ‘trade-off theory’ (which I discuss in my book on p.164). The idea is that from the perspective of the pathogen, virulence (the detrimental effect of the pathogen on the human) is quite epiphenomenal. The pathogen does not get an evolutionary advantage out of harming the host. Some of the most widely spread viruses, for instance, cause no symptoms at all in the overwhelming majority of people. My favourite, JC virus, is present in about 50–70% of all people worldwide, and generally does not cause symptoms unless the person in question is immunosuppressed (at which point, sadly, it can cause some dreadful consequences – progressive multifocal leukoencephalopathy, which occurs in HIV/AIDS and with the use of certain medications for neuroautoimmune disorders that reduce immune system activity in the brain, is caused by JC virus reactivation, and is pretty unpleasant). Ubiquitous viruses like EBV and some herpesviruses are generally very indolent, to the point of being asymptomatic in many cases. In some instances, they might even be beneficial – gammaherpesviruses are associated, at least in animal models, with an increased resistance to Y. pestis and Listeria monocytogenes.
Has smallpox been wiped out, or is it still around?
Yes. As in, it’s been wiped out, and it’s still around. And this is one of the bigger travesties of the world.
Unfortunately, humans are what they are, and the two superpowers, the US and the USSR at the time, decided that each would be allowed to keep actual smallpox samples. The last outbreak of smallpox was not ‘natural’ but a laboratory escape, killing a medical photographer, Janet Parker, in 1978. In 1984, the WHO authorised two labs – VECTOR in Koltsovo, USSR, and the CDC in Atlanta – to store samples of it. And this has ignited a debate that keeps coming up all the time.
Is the T-virus in Resident Evil a fictional virus or is there an actual disease that resembles it?
I like discussing fictional viruses. Heck, I’ve discussed one in my book – MEV-1, from Contagion –, which is supposed to be Serious Academic Literature. Fiction sometimes allows us to explore what could be, which is just as important as what is when it comes to preparing for bad things to happen.
If a person is fluent in multiple languages, what is the language of their thoughts?
When you speak multiple languages, you come to think of languages as tools. I like German, for instance, which I consider a great tool for expressing logic, and I find German absolutely wonderful for discussing mathematics. I love the conciseness of modern Hebrew, and it’s my go-to for putting something down quickly. I think Russian gets certain emotions across in ways I have a hard time doing in other languages. And so on.
You rarely use a single language for thoughts, although you do tend to default to whatever your dominant language is, which may or may not be your native language (my dominant language for thinking is English, but it was only the third or fourth language I’ve learned). Equally, I used to think primarily in Hungarian as a child (the language we spoke at home most of the time), but increasingly found I had a hard time expressing certain ideas. I learned both my first and second profession in English, and I have no idea what an outer automorphism group is in Hungarian, never mind the more specialised language I learned in law.
Is it possible for someone who speaks more than one language to identify what language someone is speaking just by hearing them speak for a few seconds (no context)? If so, how many languages can this be done with?
Yes. You can even pick this up with languages you don’t speak. For instance, I speak about a hundred words of Amharic (and about a quarter or so of those would fall into the ‘colourful language’ category), but can recognise it anytime.
One of my native languages (Hungarian) is quite peculiar to me in that way. I can recognise it even if you speak it very faintly. I may not even be able to understand individual words, but there’s a ‘rhythm’ and ‘tonality’ to it that you can understand. The best analogy is this: say you’ve listened to your fair share of modern minimalist composers. If I gave you a new Philip Glass soundtrack, you’d immediately recognise it. There are some motifs that are just ‘so Glass’. This actually goes beyond identifying the language and starts to involve the semantic layer. My wife does not speak particularly good Hungarian, but can understand quite a bit of what’s going on from tone and affect.
Can a language be replaced by another one? If so, how would this happen in an example of a real world scenario?
Yes. This happens all the time, and unfortunately it’s one of the major causes of languages becoming extinct.
By way of example, let’s consider one of my favourite languages, Yahgan, spoken by the Yaghan people in Tierra del Fuego. As languages go, it’s had pretty bad cards to begin with: it’s a language isolate, meaning it’s not related to any major language or linguistic group. In February of last year (2022), Cristina Calderon, the last native speaker of Yahgan, succumbed to complications of COVID-19 (at the ripe, old age of 93), and with her, Yahgan became effectively extinct. Which is a pity, because as languages go, it’s got some spectacularly fun features. But I’ll leave the language nerding for later.
If you had the ability to speak any language fluently, which one would you choose, and why?
My shortlist:
Arikara, a critically endangered Caddoan language. It’s spoken by fewer than a dozen people on the Fort Berthold reservation. Critically endangered, it is not mutually intelligible with any of its close Pawnee relatives.
Lower Arrernte is not actually one of the Arrernte languages, although it is an Aranda language, i.e. related to it. Brownie Doolan perrurle was the last native speaker of Lower Arrernte, and he passed away in 2011.
The Maidu languages, which are either extinct or almost so. Maidu is a language subfamily, although where it fits into the wider scheme of things isn’t all that clear. It used to be spoken in California by the Maidu peoples.
Osage, another critically endangered language – was spoken mainly in Oklahoma, and has been considered extinct since 2005, when the last native speaker, Lucille Robedeaux, passed away.
Sireniksky is a Yupik language that’s also extinct. There’s actually disagreement as to where it comes from, or whether it’s actually a Yupik language – there’s actually a theory that it’s the only known example of a third separate branch of the Eskimoan languages, along with Yupik and Inuit languages. It was displaced by Siberian Yupik, and of course Russian.
Are law firms allowed to charge as if they were doctors, despite the fact that it requires far less training and skill set than medicine does? What can be done about this issue?
I worked for a Big Law firm. To say we charged like doctors is a gross insult. Very few doctors charge remotely what we did.
There’s a quite simple reason for why that’s the case: spending seven figures (or more) with us could save a company from being sunk by a lawsuit. Think about that for a moment. We routinely acted in matters where tens or hundreds of millions were at stake, and often enough, much more. The amount of money at play in such high stakes matters is orders of magnitude beyond whatever impact an individual physician may generate. You don’t have to like this fact of life, but that doesn’t make it any less a fact of life.
If the UK were invaded, would it be legal in UK law for a civilian to murder an enemy combatant if the opportunity presented itself?
The laws of war basically separate people in two classes – combatants and non-combatants. A combatant cannot be prosecuted for the lawful conduct of hostilities which, yes, involves killing the other side’s combatants (they can, however, be prosecuted for war crimes). They are entitled to the panoply of rights that combatants have upon becoming non-combatants by way of surrender or becoming hors de combat, such as being wounded. On the other hand, a combatant can generally be killed with pretty little justification – if the kill is legal (the person is a combatant at the time, and the kill did not involve perfidy or some other form of illegality), it is what it is.
Non-combatants enjoy extensive privileges in contrast. Not only are they entitled to protection, even if they are living in occupied areas, they are also categorically not permissible targets for military action. They and their property enjoy extensive protections. They cannot generally be interned against their will unless they are prisoners of war, they cannot generally be forced to work and there are various other protections they enjoy. Most of all, they cannot be killed for no reason other than not liking the occupying power.
This gets complex where someone decides they’re going to blur the lines.
I passionately hate the meme that the UK does not have a written constitution, because, for one, it’s not true, more importantly however, it distracts from the bigger question, namely what a constitution is.
Now, a lot of how we perceive constitutions comes from the US idea that a constitution is a ‘supreme law of the land’, a superordinate legal norm against which all legislative and administrative action comes to be tested. This is not, actually, a necessary implication. Constitutions can be superordinate norms and it makes sense for them to be so. On the other hand, they don’t have to, nor is it a necessary implication of a constitution that it should be a superordinate norm. Nor do written and codified constitutions exclude reliance on other norms that happen not to be written and codified. An example would be constitutional conventions, which are basically political customs – plenty of countries with proper, codified constitutions have them. For instance, Germany has a codified constitution (even if it is not called a constitution but a Basic Law, for historical reasons), but that doesn’t provide for who gets to nominate the President of the Bundestag in the circumstances where a party has a plurality but is in opposition. It turns out that in this case, it would still be the largest party that would get to provide the President of the Bundestag, i.e. there’s no horse-trading between coalition parties to have a joint candidate for President of the Bundestag. This isn’t written down anywhere, incidentally.
Don’t American law students study the famously unwritten British Constitution? Are the details superfluous?
The British constitution is very much written.* It’s not codified. There’s a difference. For instance, nobody would argue that the Human Rights Act 1998 isn’t part of the British constitutional order, and last I checked, it’s pretty darn written. What it isn’t is codified, meaning there isn’t a single constitutional document.
Part of the reason for this is that we do not have coequal branches of government. The UK has legislative supremacy, i.e. Parliament is not fettered by the judiciary. There’s no such thing as judicial review of primary legislation (as opposed to secondary legislation aka statutory instruments, and administrative action) the way there is in the US post-Marbury v Madison. Therefore, there’s no real need — or use — for a superordinate norm against which legislation may be tested. John Griffith said “the constitution is what happens … if it works, it’s constitutional”. The lecture (and later article) in which he made this poignant remark was titled The Political Constitution, which I think is a great way of describing the UK’s constitutional order.
Is law school intrinsically worth it? Even if you don’t become a lawyer, is it worth it to go through the process of law school?
The common perception is that law school is spent studying the law. This is maybe true in the vaguest outlines. Let me give you an example: like everyone pursuing a qualifying law degree in England and Wales, I had to take criminal law. Part of that was learning the elements of certain offences. I know to this date that theft is the (1) dishonest (2) appropriation of (3) property (4) belonging to another (5) with intent to permanently deprive. Would I actually know how to defend a theft charge? Probably not. How each of these parameters are applied in practice is not a lot like what you learn in law school.
You do, however, learn a lot about reasoning. For instance, ‘dishonesty’ is an element of the offence. In England and Wales, a finding of dishonesty requires both a subjective and objective element (R v Ghosh), i.e. not only does your action have to be something others would consider dishonest, but it must also be something you yourself would consider dishonest. How would one arrive at that finding? How would one make the pleading for that? Would, say, someone believing that they are enriching themselves unjustly qualify?
If the vampire and zombie apocalypse is coming: Whitehall, call me.
So, going after all the vampires would clearly be a violation of the Equality Act, and essentially a disability-based extermination. That’s a big no-no. Assuming we consider vampires still somewhat human, it would be a human rights violation. Since death in England and Wales is generally brainstem death, it would be very difficult to conclude that a vampire, once turned, is “dead” and hence no longer deserving of human rights.
Vampires would thus be entitled to “reasonable accommodations” (such as working from home/castle/coffin or jobs that involve only night-time work), along with non-discrimination provisions. You would not, for instance, be allowed to refuse renting to the Karnstein family just because they have a predilection for blood.
Having diplomatic status is sovereignty but it’s nearly impossible to obtain so, being a sovereign citizen does exist if you can obtain diplomatic immunity?
I do love me some sovcits.
Not only is diplomatic status not sovereignty, it is the very opposite of sovereignty.
You are granted diplomatic status by presenting your letter of credence aka diplomatic credentials (this does not mean an ID but a letter from your sending country’s or institution’s authorities to your receiving country) to the receiving country’s authorities. This is a pretty big show, incidentally – typically, they are presented to the receiving country’s head of state in person.
It’s at the receiving country’s leisure to accept or reject this. A person isn’t a diplomat in a foreign country by their sending country’s fiat but by the receiving country’s consent (called agrément), which incidentally can be pretty much withdrawn at will – every country has the sovereign right to kick out any diplomat for any reason they want (Art. 9 Vienna Convention), and ban them from the country, a process generally known as PNGing (for persona non grata, and typically pronounced as ‘pinging’).
As a math major or mathematician, what makes you roll your eyes every time you hear it?
The assumption that we all do the same stuff.
I’m doing a very specific kind of applied mathematics, day in, day out. What I do for a living has very little to do with what Alon Amit, one of my favourite Quorans, or, say, Edward Frenkel does. I could, technically, go about my work without having the slightest idea of Galois theory. Of course, that would be a sad and diminished existence, but hey. On the other hand, I need to know how to use a continuous wavelet transform with a Morlet wavelet to discern the way periodicities in the incidence of infectious diseases have changed (and if you are interested in that sort of stuff, you can go buy my book). I need to be able to construct Lyapunov functions and use them to reason about the stability of equilibrium states. And I have to be able to tell whether a certain control parameter, such as vaccination rate, is going to result in complex non-linear behaviour.
As you move towards more advanced maths, it becomes less about dealing with numbers and stuff, and more about logical manipulations of various concepts. We want a convenient way to refer to certain things for which certain statements are true. To give you an example you are likely familiar with: we could say “x is a member of the set of integers greater than one that is not the product of two natural numbers”. Or, we could just say “x is [a] prime”. This is useful when you have to describe more complex things. A group, for instance, is a construct that consists of a set and an operation. A set and an operation are a group if they exhibit the group axioms — properties that are the “requirements” for a group. What these are are not so relevant here as the fact that life is a whole lot easier for us if we can simply say “group” and know that everyone else will know what we’re talking about, including, importantly, what specific features this thing we allege is a group would exhibit, from which we can then derive various other properties. For instance, one property of a group is that the operation is “closed”, meaning that performing the operation on any elements of the set will give you another element of the set. We know that addition of integers is a group (the set is the set of integers, the operation is addition). Consequently, we have proved that if x and y are integers, so will x + y be an integer. Trivial, yes, but it gets much more complicated quite quickly.
Many of these concepts are, essentially, sets of rules and conditions. Often, they’re subspecies: an Abelian group is a group that satisfies some specific criteria. A ring is a kind of Abelian group that meets some additional criteria. And so on. Mathematical terminology is intended to capture these logical distinctions in a sort of shorthand.
Now, there’s no good meningoencephalitis. They’re all bad, even the relatively treatable forms. The problem is that the brain does not have a lot of tolerance for swelling, and therefore pretty much any inflammation in the brain is going to be a massive problem.
The reason why PAM is the worst kind of meningoencephalitis is because the pathogen involved, N. fowlerii, is extremely difficult to kill. Naegleriasis is an amoebiasis rather than a bacterial infection, meaning that the usual method of solving problems – throw antibiotics at it – doesn’t really work. The most effective drug until recently used to be amphotericin B, which is notorious enough for its side effect profile to be nicknamed ‘amphoterrible’. From what I gather from an acquaintance who has had the pleasure of a personal acquaintance with the darn thing, it does its nickname justice. It’s not the most unpleasant drug in the infectious disease arsenal, because our friends in Infectious Diseases have things like literal arsenic in antifreeze (melarsoprol), but it’s pretty bad. And that’s before dealing with the problem of actually getting it to the site of the action, which often requires intrathecal administration (into the spinal canal), which has an entire set of issues on its own. Oh, and this treatment is effective in around 1.5% of cases. That’s right: case-fatality ratio for PAM is around 98.5%. Survivors are often left with permanent neurological sequelae.
What diseases are close to being cured by medical research, and why are they progressing faster than others (like diabetes)?
Probably the biggest success story is cystic fibrosis (CF). It’s gone from essentially lethal in childhood in the 1980s, to survivable into adulthood with serious interventions (such as a double lung transplant) in the 1990s to now having the “geriatric shift”: patients are getting old enough to need care for long term effects as they are now living into, if not a ripe old age, at least past previous life expectancies. In the West, with good care and management, especially in patients who have a druggable CFTR mutation that can be treated with the CFTR modulators (the ‘caftors’: lumacaftor, ivacaftor &c.), hitting 30 is considered to be more or less normal, and a child born with CF today has a better chance of seeing their children and perhaps even grandchildren than anyone with CF has ever had. And all this played out over the span of about 30–40 years.
What exactly happens during a clinical trial? How do you become a participant in a clinical trial?
See a clinical trial, and you’ve seen exactly one clinical trial. There are many different clinical trials at many different stages, for many different conditions, for many different interventions. No two clinical trials are the same – this is not an exaggeration, this is a rule: you cannot ethically do the exact same clinical trial again (the key ethical consideration for a clinical trial is therapeutic equipoise – that is, there must be a genuine question, you cannot do a trial if you’re sure that a treatment will definitely work or definitely not work)
A PET scan is basically a form of scintigraphy. The idea of a PET scan is something along these lines: a radiotracer that undergoes ß+ decay is injected, and accumulates where the tracer is taken up (this depends on the tracer and the physiology being imaged). As the tracer undergoes ß+ decay, it emits positrons (thence the PE – ‘positron emission’ – part of the name). When positrons hit electrons, they annihilate and emit a gamma photon. This is then imaged using a gamma camera, which basically measures the intensity of light,* in this case, gamma light. The result is a picture of how much gamma light is emitted from any point in space, thanks to a nifty little thing called an inverse Radon transform (that’s ruh-DOWN as opposed to ray-DOWN, which is how you are supposed to pronounce the element, to which it’s entirely unrelated).
Now, I’ve gone through this length to explain how the image is acquired because it helps us understand what the image is (which is always something you should have at the forefront in medical imaging). It’s an image of signal intensity attributable to a point in three-dimensional space (although we obviously look at 2-dimensional projections of it), where that signal intensity is the measure of the amount of ß+ decay originating from that point, which in turn is a measure of how much of the tracer is taken up by tissue at that point, which in turn is biologically meaningful (e.g. if the tracer is 18fluoro-deoxyglucose, which is a radiotagged sugar, it will show areas of higher glucose metabolism). In other words, for any point in space (and therefore, for any projection thereof), we have one value for any point in space.
What are the challenges faced by the government in financing treatment for rare diseases?
Pretty much the same as the private sector’s: they’re rare.
Drug development hinges on economies of scale. The cost of bringing a drug to market is the same, whether it treats a million people, ten or, as is the case for some drugs, a single individual (Milasen, for instance, is an oligonucleotide antisense drug developed for one individual patient with one particular mutation causing Batten’s Disease). If a finite amount of money could be used to bring a drug to market that will give 5–10 quality-adjusted years of life to hundreds of thousands, or save the lives of maybe a hundred, the calculus is quite clear.
Are there any historical examples of self-treatment or self-diagnosis that led to significant advancements in medicine or public health?
There are cells in your body called mast cells that are basically big containers of histamine. When these cells degranulate, they spew this histamine all over the place, and you experience what is known as an allergic reaction. So, it would probably be cool if we could keep these cells from blowing up, right?
It turns out that this is a relatively workable idea. A young man by the name of Roger Altounyan, who came from an Aleppo Armenian family, knew of a folk remedy widely used around the Mediterranean – khella, a weed that is boiled down to a tea and used for asthma, colic and just about anything else. The problem is, khella is tremendously unpleasant in terms of side effects. It does, however, stabilise mast cells quite well. Altounyan figured out how to create an analogue, cromoglycate, that does the same but minus the horrid side effects.
What would happen to your body if you’re exposed to 10 times the magnetic power (30T) of an MRI scanner?
Most of the weird and/or unpleasant things that tend to happen in a properly operated MRI (by which I mean nobody is letting the guy with all the spinal hardware into the scanner room) don’t actually happen due to the static magnetic field. In MRIspeak, that’s normally denoted \(B_0\). If you read MRI safety literature, you’ll see another symbol a whole lot more: \(\frac{dB}{dt}\), known to its friends and relatives as “flux change”. In this case, \(dB\) stands not for decibels but is rather the change in the total field strength. It turns out that if you have a high frequency of change (and as we know, frequency is the inverse of the period), then even relatively small fields can build up quite a bit of flux change. The MRI has a part called the RF transmit coil (or sometimes just “coil”) that transmits a much, much smaller magnetic field that, however, varies rapidly. It’s called an RF coil because the transmitted signal is in the radiofrequency range, so in the high hundreds to low thousands of MHz (megahertz). If you get a tiny field that flips a hundred million times a second, you’re going to get an awful lot of flux change. By far the biggest problem is that this field coexists, regrettably, with an electric field \(E\) (we’ve got Maxwell to thank for that). The Maxwell equation, of course, tells us that \(\del \times E = \frac{dB}{dt}\), which in short means that if you oscillate the magnetic field, you’re going to get an electric field perpendicular to the magnetic field lines that is going to oscillate at the same frequency. This electric field dumps energy into whatever is at hand, and if you’re in the Magic Donut, that energy will end up in you. There are largely two things that happen. One is the result of the fact that you’re only seventy-odd percent water. A good part of the rest are charged particles, which are accelerated by this field and slammed into the nearest water molecule. This creates vibrational energy, which is also known as heat. This is called Ohmic heating, and if you’ve ever had one of those old space heaters that were ridiculously unsafe but incredibly cosy after being drenched by the rain, that’s how they worked. The bigger issue is dielectric heating. If you vary the electric field fast enough, all sorts of polar molecules are going to try to align themselves with that field. Unfortunately, that seventy-odd percent of yours that is water? That’s a polar molecule. If you flip water molecules around their axis a couple of hundred million tiems a second, you get what everyone else knows and loves as a “microwave oven”. Higher field strengths involve higher RF frequencies (for reasons you might find out by googling Larmor frequencies), so the RF coils in a higher field will have a higher \(\frac{dB}{dt}\) and hence a higher \(\frac{dE}{dt}\). That in turns means more energy dumped as heat. It’s worth noting that this energy dumping as heat is pretty minimal. What happens a lot more often is if there’s something to be heated. There are two flavours of this. One is what we call inductive heating. This happens when the body, which is an okayish conductor, comes in touch with something that isn’t, such as the bore. That generates a resistance, and we know what happens to resistances. The other alternative is anything that creates a loop. By far the most frequent loops that generate RF burns are humans. If you are raising your arms up, don’t touch them. If you’re holding your arms down, don’t touch your hip. Don’t let cables touch your abdomen. More importantly, don’t let coils of cables be anywhere near you.
How did the scientists of the Manhattan project know it would work?
This is science. You never know if something will work. You do, however, have a degree of belief – in the statistical, not the religious sense –, and that belief tends to be the result of evidence (ideally). By the time dawn broke over Alamogordo on 16 July 1945, there was a pretty impressive amount of evidence that it would work just fine.
Much is made of Oppenheimer as an ‘administrator’, but I think that misses the point. There was a logistical-administrative genius at work at the Manhattan Project, but that was Leslie Groves, not Oppenheimer. Rather, Oppenheimer’s principal role was as a buffer.
Why would Britain have had to ask for Australian permission to test nuclear weapons in the 1950s in Australia? Britain owned Australia in the past so surely Britain can do whatever it wants in Australia?
In 1907, at the Imperial Conference, a number of self-governing British territories, such as Australia and Canada, became dominions. This made them subjects capable of independent standing under international law – Australia, for example, was one of the founding nations of the League of Nations. In 1931, the Statute of Westminster further increased the sovereignty of dominions. This basically put into effect the Balfour Declaration of 1926, which declared that the dominions were co-equal with each other and the United Kingdom. It also, implicitly, repealed the Colonial Laws Validity Act 1865 and essentially terminated the ability of Westminster to legislate for the Dominions. By the time you’re referring to, Australia has very much been an independent sovereign that in no conceivable way was “owned” by Britain. This includes participation in collective security agreements the UK was not party to, such as Australia and New Zealand’s alliance with the US – and, of course, it participated in the UN as a fully-fledged sovereign.
There was a project in the early 1980s run by the KGB called RYaN – raketno-yadernoye napadenie, roughly translating to ‘nuclear missile attack’ and sometimes wrongly spelled RYAN. The purpose of that was to watch for indicators of what in British governmentese is called ‘transition to war’ (TTW). The later Soviet leader, Yuri Andropov, was then the head of the KGB. Even by the standards of KGB functionaries, Andropov was paranoid to a clinical degree, at least partly due to his brain swimming in an increasingly potent pickle of his body’s own waste products due to untreated renal failure. More so, though, he witnessed a bunch of Soviet-trained servants of the Communist repression in Hungary get hung from lampposts during the 1956 Revolution, which left him with a life-long ‘Hungarian Complex’. Andropov genuinely believed that the West, with Reagan’s encouragement, was going to get the drop on the Soviet Union.
There’s a fear of dirty bomb used by some malicious non-governmental actor (terrorist organization). But is that fear even grounded? Is it that simple to construct a dirty bomb if compared to the “standard” nukes used by countries?
The entire edifice of control over the nuclear means of our fellow man’s destruction rests on a very simple principle: nuclear weapons need fissile material. It’s impossible to get fissile material easily (you can technically go and buy small amounts, but even that’s tricky). It’s impossible to make fissile material easily – you’ll have to master an entire industry called uranium enrichment or a different one called plutonium breeding. And there’s a minimum of how much fissile material you need to realise a yield. The ‘critical’ in ‘critical mass’ reflects this: critical mass plus one gram equals nuclear boom, critical mass minus one gram equals no nuclear boom. The effect of this scheme is that as long as nobody sells fissile material, even relatively well-funded state actors – think Iran, North Korea – have an enormously difficult time getting enough fissile material to play the nuclear game. Non-state actors pretty much stand not a snowball’s chance in hell.
Could you build a nuclear weapon off from an information online?
Assuming you have the fissile material, the answer is largely yes.
The US has, on multiple occasions, ran “Nth Country” experiments, mostly in the pre-internet era (if they did any in recent years, it is classified and will remain so for a long, long time). These have proven out that a physics PhD who did his doctorate on an area other than nuclear physics could, with the help of a decent college library (and, I assume, a friendly college librarian who has few compunctions about the weird guy asking for strange publications involving things like explosive lensing, aerogels and the properties of tantalum-182), build themselves a nuclear bomb.
How clever do you have to be to get a first at Oxford?
I’m not, by any stretch of the word, clever. I am generally a hard worker and someone who has probably well above-average stamina. Finals is a marathon, and you need to be able to both have great mental control (not let how you feel you performed one day affect how you are going into next day’s paper) and an ability to work extremely hard during the revision period.
What are your favorite places to study at Oxford University?
Hidden gem: the Codrington Library at All Souls, now known as the All Souls College Library to distance themselves from Christopher Codrington, who might have donated a lot of money to All Souls, but was also a slaveowner. The name is gone, the statue is kept.
How good is food at Oxford University? Is it varied among different colleges?
It’s quite decent – and yes, it varies as food is made & served by the colleges. It’s British(-made) food, so take that with a pinch of salt (literally – it will taste super salty, but actually be devoid of actual salt, because the government replaced all salt with potassium chloride). Food at Formal Hall is generally very nice, food at lunch and informal hall ranges from good-quality institutional food to quite nice, actually. The typical staples are the usual English breakfast (baked beans, sausages, bacon, you get where I’m headed), whatever can be made at institutional scale for lunch, and the same again for dinner. If you go to informal hall every day for three meals at the same college, you will very rapidly grow bored of it. It’s not necessarily unhealthy, and it’s quite affordable, but it ain’t Heston Blumenthal. The same bubble & squeak will get a little dull after you’ve had it for the fifth time that term. This, incidentally, is a reason for Fifth Week Blues (incipient depression around middle of term): you realise that you have five more weeks of every bit of meat getting drenched in the same made-from-powder gravy.
How do undergraduate law students get grades at Oxford? Do they have quizzes? Do they write essays or what for the finals? Are there any finals or mid-term exams at all?
Law at Oxford is a pretty… interesting subject when it comes to grading. There you are, after three or four years of suff… I mean, education, and all of it will boil down to a set of papers in one week. That’s it, that’s the whole thing. Nine three-hour papers. Nothing else matters.
Over the past years, you’ve sat collections (like term exams) at the start of every term. You also sat Moderations (Mods) in your first year, which comprised three papers, in subjects you won’t be tested on in finals. But none of this matters, because it’s a brand new day, and everything starts anew. It’s hard to overstate the psychological pressure inherent in this. We all have good days and bad days, but you absolutely cannot have a bad finals week. This is an absolutely gruelling system that rewards not just knowledge and excellence but also the mental stability to perform no matter what. Training is what you default to when all else fails, and this is training at its harshest.
Does an Oxford degree provide conclusive evidence that one person is more intelligent than another?
There are relatively few people around who have done as well on their undergraduate degree at Oxford as I did. The number would probably be, if I counted everyone alive and assumed that there are the same number of equivalents from the Other Place, in the three digits, and that’s assuming every year’s top 1st in every subject did at least as well as I did or better. Off a sample of eight billion, that’s a pretty low number. There are more billionaires alive than people with a top 1st from Oxford.
There are probably tens of millions, if not more, people who are vastly more intelligent than I am. Some have top degrees from other universities. Some don’t have any degrees at all, or yet. Some could never get into Oxford and many, if they did, might not do well. There are many things involved in getting in, and doing well at, Oxford. Intelligence is at best part of the story. If I had a limited number of skill points to put into various skills to ensure success at Oxford, intelligence wouldn’t be my first or even second choice. More has been determined by industry, hard work and sheer bloody-minded determination than intelligence. Evolutionary biology calls this the Anna Karenina effect: no one factor ensures success, but rather the presence of all of them is necessary (which is why e.g. zebras were never domesticated like horses - they lack one crucial factor, namely the right temperament).
People with IQs below 83 are unemployable in western society, according to Jordan Peterson. Even people with IQs between 85-100 are mostly confined to low paying jobs (less than 2x the minimum wage). Why is this issue ignored by society?
Unlike Dr Peterson, I graduated with a top 1st from the University of Oxford, and got my graduate education there, too. Without being so crass as to discuss what I make, it’s… well, it’s rather more than twice the minimum wage. I managed to do so with an IQ of 83–87, depending on test (all proper testing, administered by actual people, not some online stuff). So I suppose I might have an opinion on the matter.
The idea that people with an IQ of 83 are ‘unemployable’ is ridiculous. There probably is some cut-off at which point it becomes difficult to employ someone without constant supervision, but people in the low 80s and even in the 70s can work just fine.
Can money really buy happiness? If yes, then how and if not, then what can bring happiness?
Money can’t buy happiness, but it can definitely buy the things you need to do what makes you happy, including – to some extent – time. It’s a common misunderstanding that money can’t buy time. That’s only the case if you manage it badly. If you can accrue money faster than what it costs you to buy an hour of your time, then this is absolutely no issue. 90% of what you do in your waking hours are things you can make someone else do, and usually for pretty nominal cash. I think there’s something good about doing many of these activities yourself – for instance, I’d never hire a personal chef, as making food relaxes me and it’s a good way to keep whatever motor function I’ve got left in my hands mostly in practice. On the other hand, I hate doing the dishes, and if I had to fly to Sotheby’s in London to bid on the last dishwasher in existence, I’d raid my savings to make sure to get it. There are very few things money cannot buy (love, loyalty, friendship are some examples). For all else, you just need to know how to spend your money to buy what it is you really want. If there’s one thing I’ve seen affluent people spend differently on, it’s that vastly more of their spend goes towards things that either buy them time, or buy them experiences to fill that time with.
For starters: invoking diplomatic immunity is not really optional. The immunity does not belong to you as a person, it belongs to the state that sent you or your family member. As such, many interactions that you might be able to iron out pretty easily as a normal person will become pretty… complicated.
What do drunk scientists discuss more at cocktail parties: The multiverse, wormholes, alien sightings or time travel?
We are, in fact, boring.
Neither.
You’d be surprised by what passes for normal conversation among scientists at cocktail parties. A cocktail party is where you don’t talk shop. If you do, you are marking yourself as a terrible bore, and you won’t be invited ever again. There’s nothing more boring than someone holding forth about abstract nonsense at a cocktail party.
