Share this page!

Philippe Lemoine Profile picture
Twitter logo
11 Dec, 5 tweets, 4 min read
Beautiful. (h/t @Scientific_Bird) ImageImageImageImage
Now, by the Law of Lemoine about papers that purport to show the world is terribly unfair, if I go and dig into the paper's data/tables, I'm almost guaranteed to find something politically incorrect. I should work but I don't know if I'll resist the temptation.
Okay I couldn't resist and I didn't have to look for very long before the Law of Lemoine was once again vindicated 😂 papers.ssrn.com/sol3/papers.cf… ImageImageImage
The part I really love is that when they reminded the programmers that discrimination could make their algorithms biased, it actually *increased* bias, because they were already biased in favor of women so it made them even more so. This paper is truly a narrative-killer 🤣 ImageImageImageImage
To be clear, I think it's very likely this effect isn't real because as a rule priming effects aren't, but I'm going to ignore that and "believe science" for once because it's so beautiful 😄

• • •

Missing some Tweet in this thread? You can try to force a refresh
　

Keep Current with Philippe Lemoine

Philippe Lemoine Profile picture

Stay in touch and get notified when new unrolls are available from this author!

Read all threads

This Thread may be Removed Anytime!

PDF

Twitter may remove this content at anytime! Save it as PDF for later use!

Try unrolling a thread yourself!

how to unroll video
  1. Follow @ThreadReaderApp to mention us!

  2. From a Twitter thread mention us with a keyword "unroll"
@threadreaderapp unroll

Practice here first or read more on our help page!

More from @phl43

Philippe Lemoine Profile picture
13 Dec
What very intelligent people call "science" is indistinguishable from magical thinking. Image
By the way, I see that all the pro-lockdown hysterics are outraged that Johnson is still listening to crackpots like Gupta, but what they're effectively arguing is that instead they listen to the geniuses from Imperial College, as if their track-record were any better.
Not only are the researchers at ICL who were arguing that a 2-week long lockdown in September would magically make later restrictions unnecessary are evidently incompetent, but they're also intellectually dishonest, as I showed in the post below.
Read 6 tweets
Philippe Lemoine Profile picture
5 Dec
The reception of this post has been very positive, thanks a lot to everyone for your kind words and for sharing it, but I also got some of the same criticisms I get every time I publish something, so I wanted to address them quickly. 1/n
So one complaint I get almost every time is that, if I think I'm right, then I should try to get my work published in a peer reviewed journal. The suggestion is often that, if I don't, it's because I'm afraid it wouldn't stand up to scrutiny. 2/n
There are many reasons why I generally don't want to submit my work to peer reviewed journals, some of which I discuss below, but let me start by saying that my fear of the towering intellects who run Nature and saw no problems with Flaxman et al.'s paper is not one of them. 3/n
Read 23 tweets
Philippe Lemoine Profile picture
4 Dec
Okay, I know most people aren't going to read this because it's long (though you should!), so let me go over the main points quickly. 1/n
It's a critique of Flaxman et al.'s paper, which is constantly cited as proof that lockdowns are the only interventions that really works. It's already been cited almost 450 times even though it was only published in June. 2/n nature.com/articles/s4158…
It found that only lockdowns really had a meaningful effect on transmission in Europe during the first wave. Here is a chart that show the results in a few countries that were included in the study. 3/n
Read 34 tweets
Philippe Lemoine Profile picture
1 Dec
Unfortunately, this piece will only strengthen the myth that China engaged in massive and systematic fudging of COVID-19 data, when if you read it carefully it shows no such thing and doesn't tell us anything important that we didn't already know. 1/n
In particular, when the tweet says that China underreported COVID-19 numbers, what the article shows is just that, as there weren't enough tests, the authorities at times only reported cases that had been laboratory-confirmed by PCR but not cases identified by symptoms. 2/n
This is not new, I already discussed this at length in my essay back in September. We have known for *months* that the definition of a case used by the authorities changed several times, which affected the numbers. 3/n quillette.com/2020/09/06/the…
Read 14 tweets
Philippe Lemoine Profile picture
18 Nov
Wow, I hadn't actually read the Nature paper that allegedly showed that lockdowns had saved more than 3 million lives in Europe last Spring, but now that I have I'm utterly shocked this worthless piece of garbage was published. nature.com/articles/s4158…
Also, something didn't make sense about the results and the only explanation I could think of implied that the conclusion people drew from that study was totally unwarranted, but it was impossible to tell from their description of the results whether my hypothesis was correct.
So I downloaded the code and ran the analysis myself so I could take a closer look at the results and, surprise, it confirmed that my hypothesis was right, which presumably is why they neglected to describe this particular result in the paper or the supplementary materials...
Read 4 tweets
Philippe Lemoine Profile picture
17 Nov
Je suis contre le confinement, mais la baisse du nombre de morts en Suède est un artefact du délai d'enregistrement des morts. Compte tenu de l'évolution du nombre de cas, ça va inévitablement augmenter rapidement dans les jours qui viennent, il n'y a pas de magie. 1/n
D'ailleurs, quand on fait un simple ajustement pour tenir compte du délai dans l'enregistrement des morts, on voit très clairement qu'en réalité ça augmente rapidement. Encore une fois, je suis contre le confinement, mais il ne faut pas se raconter d'histoire. 2/n
De la même façon, le nombre de morts par million peut sembler faible, mais c'est le nombre par jour. En France, si on prend le nombre de morts cumulés sur l'année, ça va représenter au moins 10% de la mortalité normale à la fin de l'année. Ce n'est quand même pas rien. 3/n
Read 5 tweets

Did Thread Reader help you today?

Support us! We are indie developers!

This site is made by just two indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member ($3/month or $30/year) and get exclusive features!

Become Premium

Too expensive? Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal Become our Patreon

Thank you for your support!

Follow Us on Twitter!

Like this author's thread?

Get emails when @phl43 has new unrolls!

Check out other threads from @phl43!

Read all threads by @phl43