Do you think there is extraterrestrial life, which is smarter than humankind?

That's obviously a very hard question, but luckily some "smart" people answered it for us:


On a more serious note, that equation is indeed pretty worthless. I have two related chains of reasoning to answer this question, both probably existing in some more fancy manner, but from what I know also very often wrongly interpreted.

The first reasoning deals with the concept of "life". One of the issues here, is that it's extremely hard to "create life" out of nothing. Biotechnology these days is so far evolved that it's theoretically possible to create all building blocks for life in a lab. There is just no way to actually bring them to life, even at the lowest level. It is still impossible to create a living cell out of non-living material and as far as I'm aware, they'll be stuck at that stage for a pretty long time still.
So when it comes to life, on earth or else in the universe, how did it start? If we dismiss a Godly being or an outside intelligence as an option for now (because that would simply shift the question to who created those or how they came into existence), there still is no clear answer, let alone a replicable one.
Hence, the probability that life appears by chance is extremely low. But how low? It is utterly impossible to guess this, which brings me to the second chain of reasoning.

If we talk about the existence of intelligent life, we inevitable end up with the Poisson distribution, and that's a problem.
So, let me tell you about the Poisson distribution. The easiest way to imagine it is from a binomial distribution. Suppose you have an outcome where lots of possible items could be observed, but each of them with a very small probability.
Example 1: you are in a factory that produces a million items per day. The probability of an item being broken is about 0.002% (2 per hundred thousand). The number of broken items per day is in essence a binomial distribution, but you can estimate it directly from a Poisson distribution as well, with a single parameter that is the expected value, in this case 20 (1 million multiplied by 0.002%).
In example 1 you don't need the Poisson distribution, because you supposedly know your "n" (number of items) and your "p" (probability of a "succes", being broken is a success here)
Example 2: how many different people will enter a store in an hour? This is Poisson distributed, but your n and p are unknown, despite existing. You could say that "n" is everyone in the city, or everyone in the country, or everyone passing by in that street that day (which by itself is Poisson distributed...). Your "p" is then the probability that such a person enters the store. Since it's impossible to estimate n and p seperately, you can instead estimate it by using the expected value, which is conveniently the parameter of your Poisson distribution. This could be done by observation, e.g. by counting the number of people entering the store for a few hours.

Note that in both examples, I left out some assumptions like independence, which are technically needed.
However, in summary, the number of events happening in a certain timeframe, the number of items in a certain volume or space, those are typically Poisson distributed.

So, what does the Poisson distribution have to do with extraterrestrial life? Everything!
At every step of the reasoning, we are talking about an extremely high number of possible events, but all with extremely low probability!
* How often does life get created in a billion years? The probability of it happening, as mentioned above, is extremely low. But the number of opportunities for it to happen, on Earth alone, are huge. There's a billion years of time for it to happen right!
* How many planets are there for life to be created out of nothing? The universe is pretty big... But how likely is it that this existing life would become intelligent?
* How many would broadcast at one point? But how many at a time that we are listening and at a frequency that we can observe?

I could go on and on, but the problem here is, any estimate for any of these numbers is in essence either wrong or extremely imprecise (with the exception on the number of inhabitable planets, for which reasonable guesses exist now).
Now, what happens if you multiple two things that are completely wrong? The result is much much much more wrong! (Example. If the reality is 10 and 10, then the product is 100. If I guessed it wrongly as 1 and 20, then the product is 20. My initial mistake was 9 and 10, my final mistake is 80! And that's when the error is in opposite direction and still cancelling out a little bit... 1*1=1 is 99 off, 20*20=400 is 300 off!)

So, Poisson distribution. The number of civilisations with extraterrestrial life is Poisson distributed. And the only way to estimate a Poisson distribution is observing over a certain time or space, or combination of those. We have two observations for two different Poisson distributions here:
* The existence of life: We know that life did get created somehow at least once (on Earth) over a timeframe of a few billion years and within the space of somewhere between one planet and one solar system. (Unclear if we can claim there is no other life on other planets in our solar system, but for your question extremely unlikely that it would be more intelligent than humankind)
* We have 0 verified observations of extraterrestrial life.
However, that sample from which we observe that 0 is extremely small, both in space (we can't see properly very far when it comes to noticing extraterrestrial life) and in time (we have only been observing a tiny fraction of the time, less than 100 years of a universe of a few billion years old).

So in essence, all that theory, all that literature, just dismiss it. It's just philosophy. The truth is: we know nothing and all their estimates are bad.
And I personally do agree with Randall Munroe, the creator of the comics linked above and below (xkcd.com), us searching for extraterrestrial life the way we do now, is pretty much useless.


Edited 10/26/2022 19:16:47

thoughts on the new modding update?

Edited 10/29/2022 16:20:32

What are the best places to visit in Belgium?

Friet or patat ?

How many accounts do you have?

thoughts on the new modding update?

First a disclaimer: I rarely, if ever, play mods as they tend to crash or make my computer freeze for one reason or another. This probably has to do with my computer and not with mods themselves. Similarly, idle works bad on my computer and perfect on my cell phone.

I have followed modding from a distance and I love how it can be a solution for many requests of players on uservoice that were previously ignored, or simply impossible to implement without breaking the way that standard play works.
The new modding update takes that one step further, so in that regard I appreciate it and from what I read most of the modding community as well. Warzone is slowly developing more and more from a closed game with standard options provided by its owner only (Fizzer) to an open game where users themselves can implement features. Years ago, when API's were first introduced, that felt like a big step as it allowed for tournaments and ladders organised outside the websites. Now, we are several iterations further and I can't wait to see where this will lead us all.

From a practical point of view, I was curious already and your question was the perfect excuse for me to try the tanks mod. Guess what happened... So it does look like I won't be playing modded games for a while longer, which is too bad.

Who is your favourite philosopher and which ones were the most influential to you personally?

What are the best places to visit in Belgium?

I'm obviously biased, and no single answer can be given here. It does depend a litte bit on what you like for your visit: historic buildings and museums, parties, sun and beaches, walking in nature, sports or sporting events, ...

Belgium has a more populated northern part, Flanders, where the native language is Dutch, a bilingual capital, Brussels, and a more sparsely populated southern part, Wallonia. The places to visit are very different in these three parts.

For walking or nature, you generally don't have to be in Flanders, although there are some relatively small beautiful nature reserves. Flanders is quite cycle-friendly, although not exceptionally so, with walking and cycling routes throughout. Wallonia on the other hand has the Ardennes, which is a forested hilly region that ideal for leisurely hiking.

Given our location, we are not ideal for a beach vacation either, although the limited strip of beach that we do have is very much catered towards tourists (mainly from nearby locations, Belgium itself, the Netherland, France and some Germans).

However, if you are interested in city trips and history, Belgium has a lot to offer.
In order

Number 1 to visit Brugge (Bruges) is a stunningly beautiful historical city. Most of its older buildings are protected, meaning parts of the inner city look several hundred years old, and not in a bad way! There are also several museums to visit and on a sunny day, the coast is nearby.

Number 2 to visit is Antwerpen (Antwerp) is the capital of the world (at least according to its inhabitants). That said, it's an incredibly vibrant city with lots of culture, architecture, night life and shopping opportunities. Basically, while it's not as historical as Bruges, it is considerably bigger and more diverse in what it has to offer. Next to architecture and historical building, it has some top quality museums, and was a trendsetter in the European gay community for years. Next to all the classic shopping, Antwerp also has some speciality shops and hosts the physical stores of some world famous designers originating from its fashion academy.

Number 3 to visit is Brussel (Brussels, Bruxelles), the capital of Belgium. As a capital city, it has some advantages over Antwerp (and Bruges), although personally I do feel that some of the negatives of being a capital (less clean, more poverty, more tourist traps, more busy) outweigh the advantages. Visiting Brussels is never a bad choice, although it shouldn't be the default choice for a city to visit in Belgium that it seems to be for some people.

Tied for number 4 and 5 are Leuven (Louvain) and Gent (Ghent). Two cities who are very different, but also have quite some things in common. Both are in a sense poor man versions of Bruges in different ways. Both have many historical buildings and some interesting museums. Both are slightly bigger than Bruges as well, and both are university cities, hosts of the two largest universities of Belgium (I'm a graduate of both). While Leuven as a city, is very much influenced by the university and feels very quiet, even a bit desolate, during the summer when there are no students, Ghent is much more of a city by itself that is supplemented by the unversity. I personally prefer Ghent over Leuven, but this is subjective.

In Wallonia, there are no cities that I would specifically advice for city trip, and I'd tell you to stay in a smaller city or village instead, which especially in the Ardennes region are quite charming.

Finally, some sport events to visit:
* If you are a gymnastic fan, the world championship gymnastics will be organised in Antwerp in 2023!
* If you are a Formula 1 fan, visiting the race in Spa-Franchorchamps is truly a good idea. An epic historical track smack in the middle of the Ardennes that is at risk of disappearing from the calendar every year these days. Worth a visit if you can! When there are no races going, it is possible to drive on a part of the track.
* For cycling fans, visiting Belgium is always nice, as there are several famous cycling races every year and, as was seen at the world championship in Leuven last year, Belgium is crazy about cycling and has passionate fans.
* The memorial Van Damme in Brussels is a very nice athletics event worth a visit for the passionate athletics fan
* While soccer is very popular in Belgium, I personally see no added value to watching a team in Belgium compared to neighbouring countries.
* Belgium is among the world's best in field hockey. Worth a visit if you are fan, especially since games are often organised near Antwerp.
* I believe the world championship breakdance will take place in Belgium in 2023 as well.

Friet or patat ?

Friet, or more exactly, neither actually.

For people who don't know what this discussion is about: these are two names in Dutch language for a dish that is very typical in Belgium, but also eaten elsewhere, that of baked slices of potato.
In English, this is typically refered to as "fries" or "French fries", the word French often hotly contested by Belgians who insist on calling it "Belgian fries" in English (although I'm not sure if their original is truly Belgian).

In Dutch, there are two words for them, however in Flanders everyone except Dutch immigrants uses the "friet" version. To be correct, that should be "frieten", which is the multiple, or "frietjes" which is the multiple of its diminutive. No Flemish person in their right mind would call the dish "friet" as this is the word for a single item instead, not the whole dish (which I don't expect JK to know as he is Dutch, more on that below).
Friet would technically translate to "a fry", while frieten would translate to "fries".

In the Netherlands, it's more of a mixed bag, with some people saying "friet" using the singular form for the dish called "frieten" in Flanders, and other people using the term "patat". The latter is confusing to Belgians and will result in ridicule when using in parts of Belgium and especially Antwerp as "patat" is the dialect word used for an actual potato in Belgium. Even worse, in some regions in Belgium, it's asking for a smack in the face or otherwise getting a hit, as an onomatopoeia.
While "patat" officially translates to "fries" as well, ethymologically, you can see that the world "potato" is closer to it and would be the more sensible translation.

So JK's question originates from the fact that in the Netherlands, this term is hotly contested with different people insisting on the term "friet" vs. "patat". In Belgium, this is a non-issues, in Flanders (Dutch speaking part of Belgium), everyone calls it "frieten", and in Wallonia (French speaking part of Belgium), it's frites.

Dutchies who say "Patat" are wrong, but not willing to acknowledge it @jk

What's the most elegant tool in your field?

Is Heath Ledgers performance as “Joker” in the film “The Dark Knight” the best performance ever performed on the big screen?

Have you got more interesting meetings recently?

How many accounts do you have?

So, I've broken my head about this for a bit and turns out that I really don't know...

Active accounts: 1 (this one)
Accounts that I have played competitively with for more than a few games: 2 (this one + my alt)
Accounts that I logged into in the last month(s): 1

After that, it gets murky. I made at least 2 more accounts at one point in the distant past, but possibly as much as 4, but I forgot about most of these and for any after the first two, I wouln't remember the emailaddress, nor the password, nor the account name, I only vaguely remember that I was going to make them, or actually did, but the rest is lost in the mists of time.

Then, there are multiple accounts that I may have, or could technically have access to if I wanted. My partner made an account in an equally distant past that she hasn't used for years and of which I should have the login credentials somewhere if necessary.
At least 2 people, but possibly as much as 4, did share their login information with me after they retired because they were still in a tournament together with me in the same team. However, given that for each of those cases, we are talking at least 5 years ago, probably longer, I forgot all of those as well. With 1 exception (ragingpikey), I even forgot who they were. Technically, I could get that information back if I'd check all my private messages of all my games, which I'll obviously not do.

Finally, I got temporary credentials to 2 accounts when people were on vacation, neither of which were ever used and both of which I shouldn't have access to anymore.

So, the answer is between 1 and 11, depending on your criterion, although I personally only consider this account and the alt in my clan as my only true accounts.

Dutchies who say "Patat" are wrong, but not willing to acknowledge it @jk

I blame my parents for rising me wrong...

Who is majortom85?

Which part of pure mathematics do you dislike? Isn't statistics based on derivations from pure mathematics in a sense? And hasn't pure mathematics lead to applications in fields across the board, given enough time? Multi-dimensional mathematics was once considered purely theoretical, yet finds its applications in statistics and a ton of other fields.

Who is your favourite philosopher and which ones were the most influential to you personally?

Somehow it feels like this is the question that should've received the longest answer in this topic, and it may get the shortest.

I don't have a favourite philosopher really, and no philosophers were truly influential to me.
Without giving specific names, I guess the old Greek ones, who were scientists as well as philosophers, had indirectly the most impact as they are, among others, the founders of modern science.

However, when it comes to thinking, or free thinking, I tend to try to reason myself using the data at hand, and make up my own mind, not necessarily following the lead of ideas of others all that much (which can be very frustrating, as friends and family can get annoyed that I can't just "accept" something and prefer, to a reasonable extend, to understand why/how/when/...

Finally, favourite doesn't need to mean "most influential". Some philosophers are more interesting than others simply because of their life story. So I think Galileo Galilei would be my favourite, simply due to his disagreements with the pope and spending years imprisoned for advancing the truth.

What's the most elegant tool in your field?

The word "tool" is a bit of a weird one here. I'm a bit torn between two different ones, that are both incredible useful and elegant in my opinion for different reasons.

* The Central Limit Theorem (CLT) is an absolute swiss knife in data analysis. While its technical use and definition is quite limited, in practice, it is basically underlying about everything when it comes to distributional assumptions.
Data can come from so many sources and distributions that it's impossible to plan ahead. However, all of that doesn't matter as long as you collect enough data. Because thanks to the central limit theorem, your mean is going to be approximately normal.
People only know it from the classical mean, but in modelling, it's all about the (conditional) mean, potentially behind some link function or through some construct. There too, surprisingly, it's still all about the conditional mean and p-values and confidence intervals will use certain distributional assumptions that are in essence normality assumptions at a different level.

So, personally, I don't care much about deviations from distributional assumptions up to a certain point when I'm helping people with their data-analysis. Sure, non-normality ruins prediction intervals, but as long as the point estimators are fine, you got enough data and your distribution isn't too horrible, boom thanks to CLT your confidence intervals are close enough and will have approximately 95% coverage, and your p-value is a good enough approximation for all practical purposes (and truly when it's very close to 5%, you want to be careful anyway, assumptions be damned).

* Flexible modelling, better know as additive modeling, or generalised additive modeling (GAM) through the use of p-slines or related, is extremely elegant. The CLT above can fix all your distributional assumptions, which is great, but most models still have a linearity assumption as well, which doesn't always make sense. A GAM solves that, replacing the linearity assumption with a semi-parametric approach that fits the curve on the data, and because it's actually modelling, and not some sort of nearest neighbour approach (like LOESS), you can do inference on it as well (confidence intervals, p-values etc.)
Talking about an actual tool, the statistical program R is great, and within it, the mgcv package to do GAM modelling is truly one of my favourites.

You used to be the only person to be in top 10 of 1v1 ladder in every calendar year of its existence as well as in top 100 of every season of the seasonal ladder. That must've been quite some achievement. How did you feel when your streaks ended and how happened it?