731–740 of 1001 posts
Some highlights of Slate Star Codex over the years (I'm missing archives for two of these, and I'd appreciate anyone linking to archives if you have them): Mediations on Moloch: https://web.archive.org/web/20200516181754/https://slatestar... I Can Tolerate Anything Except the Outgroup: https://web.archive.org/web/20200516181754/https://slatestar... Proving too Much: https://slatestarcodex.com/2013/04/13/proving-too-m…
Proving Too Much: https://archive.md/M1XHf
Earlier quoted context omitted.
I think that decision was worse than a sign that they have given in to market forces. They have given in to internal activists who have no desire to learn, think, or report the truth, merely use the paper as a weapon for social change.
That's Matt Taiabbis proposition too. https://taibbi.substack.com/p/the-news-media-is-destroying-i...
By the way, the Post Gazette seems to have stolen that article, so maybe don't drive business their way. https://twitter.com/mtaibbi/status/1274747185507246081?s=20
Instead, here is Taibbi's substack. If you're off a mind, a subscription is only $40 per year, but the stolen article is nevertheless free. https://taibbi.substack.com/
I stopped trusting reporters about 35 years ago, when I personally witnessed an occurrence, where a reporter was at the scene, and later read what he had written about it in the newspaper. It had practically nothing to do with what really happened, but was written in a way that most of their readers would most likely expect and endorse. I was still very young then, but it opened my eyes, and from then on, I mostly st…
The coverage of COVID has been more distressing than the actual disease. (Hint: epidemiologists specialize in studying the spread of diseases after the fact; asking them about an ongoing epidemic is like asking an expert on Roman architecture to build an office building.)
Here’s some example nonsense in the news. I regularly see the same story argue both points in each bullet:
- Sweden has had too many cases (deaths), and too few cases (people with antibodies).
- Last week 5% of 1000 confirmed cases died; this week 0.5% of 10,000 estimated cases died. “Experts” “baffled” but we are winning and can reopen (the same publication will flip the numbers and conclusion tomorrow)
- Antibodies might not lead to immunity, but the vaccine (which does nothing but cause your body to create antibodies) will be a panacea.
Bonus gem from yesterday:
New study shows kids don’t spread COVID. The numbers are based on studies of areas where schools and daycares were closed, and the kids were quarantined. Adults were more likely to catch COVID at work than from their quarantined kids. There was one (just one) school where the kids spread COVID amongst themselves, but that was probably an outlier. No one really understands how it happened.
Note that the high-level conclusion of this last article was probably right: kids usually don’t get symptoms, and asymptomatic people are less likely to cough / spread it. A stopped clock is right twice a day, I guess.
Earlier quoted context omitted.
I'm in the field - though not as prominent as Yann (who has been very nice and helpful in my few interactions with him) - and your interpretation is off. People are disagreeing with his stance that researchers should not bother exploring bias implications of their research. (He says this is because bias is a problem of data - and therefore we should focus on building cool models and let production engineers worry abo…
> He works for Facebook. He's paid with Facebook money. So why draw this imaginary line between research and production? He is paid to do research that will go into production. This is a silly standard to uphold. The sizable bulk of American academic researchers are at least partially funded by grants made from the US federal budget. If you were to enforce your standards consistently, then all of those researchers wo…
Yann LeCun isn't receiving a partial research grant from Facebook. He's literally an employee of Facebook. His job title is "VP & Chief AI Scientist" (at least according to LinkedIn).
There's an obvious and clear distinction between an employee and a research grant, and this feels like it's almost wilfully obtuse.
Earlier quoted context omitted.
If you are a history nerd, reading what people wrote 90 years ago you will realize that we are exactly the same species, and our attitudes have not changed a bit. One of my favorite readings are the essays of french philosopher Simone Weil after two visits to germany in the thirties. She was concerned with the rise of the nazis, while at the same time describing the natural and understandable forces that were making…
I've been heavily downvoted for suggesting that we're witnessing something dangerous. It's not that I think I can predict the future, but even if it continues as-is we're witnessing a loss of fauna, and this could easily mutate to something truly terrible in the next decade. The fact is Scott Alexander was my canary. If a compassionate, liberal-minded intellectual that carefully understands both sides of every issue…
By that measure, it hasn't been safe since he started his blog. In the article, he explains why he's always used a name he hopes people cannot trace to him.
"Cancel culture" is worth discussing, but it's not the topic of the article. He gives one example of it in a long list of things peoplr have done when they dislike him for the blog and discovered his identity.
Earlier quoted context omitted.
If you are a history nerd, reading what people wrote 90 years ago you will realize that we are exactly the same species, and our attitudes have not changed a bit. One of my favorite readings are the essays of french philosopher Simone Weil after two visits to germany in the thirties. She was concerned with the rise of the nazis, while at the same time describing the natural and understandable forces that were making…
I've been heavily downvoted for suggesting that we're witnessing something dangerous. It's not that I think I can predict the future, but even if it continues as-is we're witnessing a loss of fauna, and this could easily mutate to something truly terrible in the next decade. The fact is Scott Alexander was my canary. If a compassionate, liberal-minded intellectual that carefully understands both sides of every issue…
Earlier quoted context omitted.
> The implication being that the NYT wants to use real names to drive clicks and appease advertisers? This shows a lack of how journalism works. Using real names isn't to "drive clicks" and "appease advertisers." It's to add credibility to a story. Think about it: Does a furniture business advertising in the local paper care whether the victim of a shooting is named in a piece? Sure, the owner might know the victim,…
It is a pretty damn dubious measure of credibility and it has already failed when suggested for civility. I am honestly starting to think real name policies are just about hating anominity at this point.
However, if your information is revealed, don't be shocked when someone approaches you with that information because you didn't cover your tracks.
Earlier quoted context omitted.
Disengaging based on an N=1 is not the type of individual action that improves our society. Our society is built on individual (and collective) attempts to improve, and you putting your money towards journalists you found to be doing a good job is the way we leave the world better than we found it. I hope you'll consider this, because our society cannot function without quality investigative reporting. I of course ag…
> Disengaging based on an N=1 is not the type of individual action that improves our society. Statistical significance is not the only epistemological tool around. I would even argue that, outside of some scientific fields, it is really not that important (and might even lead to a lot of wrong conclusions in the context that is used nowadays, but that is a different discussion altogether). We are not some dumb statis…
This was in a piece specifically about the blog that was going to be positive, which makes this all the more inexplicable.