Steve Schaffner @[email protected]

@baym @ClementYChow I've been stopped in the last couple of years for not having my little tube of toothpaste out in the ziplock bag.

@ClementYChow Yeah, haven’t for years

My favorite part of air travel is the gentle gaslighting of the rules being enforced differently every time and the agents insisting that they’ve always been that way

@dileeplearning @DrYohanJohn @WiringTheBrain You mean, why would anyone want to do science?

@DrYohanJohn @WiringTheBrain Why would anyone want to model a phenomenon? A phenomenon is almost always a side effect of something deeper.

There is a huge difference between modeling 𝘥𝘢𝘵𝘢 and modeling a 𝘱𝘩𝘦𝘯𝘰𝘮𝘦𝘯𝘰𝘯. The success of AI/ML means that the former is eclipsing the latter in neuroscience, and I imagine in many other fields.

@baym Do you have one from a Windows phone?

For years I’ve had a photo project of windows crashes in hilarious places, and I feel like today was just a phenomenal birthday present from the universe For example, an MBTA turnstile from 2010:

@vsbuffalo Or, if your banker's uncertainty is wrong, you go broke. If you're the bank, you get bailed out by the government.

I think it's fair to say the blame doesn't fall entirely on that model, but other (e.g. regulatory) failures too. From an epistemic perspective finance is fascinating because holding wrong beliefs is tangibly costly. If your uncertainty calculations are wrong, you go broke.

I re-read this article last night, about how the Gaussian copula function is just a really bad way to model correlated risk in CDOs, and how this played a role in the 2008 financial crisis. Perhaps the highest monetary cost of a model being wrong ever? wired.com/2009/02/wp-qua…

@ksamuk @ZaminIqbal Thanks for advertising this. I ran into this last year and it's infuriating. Why they still claim to be outputting VCFs is baffling.

PSA, since I've now seen problems arising from this: As of Sept 2023, GATK no longer makes missing genotypes explicit in VCFs (i.e. "./." as in the VCF spec). Missing genotypes are instead coded as "0/0" with a DP=0 format field. See attached for examples.

@EdHoopman @LogosMMXX @BajaGorgeGirl @celinegounder @Laurie_Garrett The last studies I've seen have been quite convincing that rebound is more common after paxlovid. It absolutely should still be prescribed more, however.

@LogosMMXX @BajaGorgeGirl @celinegounder @Laurie_Garrett You can get rebound just as easily without taking paxlovid. Look it up

1/ Lots of my friends are getting COVID right now... & they're having trouble getting Paxlovid. (@Laurie_Garrett, feel better soon.)

@johnbind2 @DegenRolf I think you just got it

Reputable brain journal claims that a non-invasive form of brain stimulation can unleash psi powers in humans. sciencedirect.com/science/articl…

@jcgallaher @RNSuperHero I probably shouldn't have read this 12 hours after hernia surgery.

I don’t understand how shared rooms are even allowable with HIPAA laws. Because I guarantee bed 1 can hear everything you’re telling bed 2.

@vsbuffalo I don't think you need any actual training in philosophy to recognize that as nonsense, besides being wildly inaccurate as a summary of the field.

Hume must be rolling over in his grave. When I read this, I think scientists need more training in philosophy. The complete blurring between is and ought, between a positive statement like “genetic variability exists” and values “is it good or bad” is poor reasoning in my view.

@baym It should be enshrined right next to 'This is our fucking city'.

“Together we can not hit a bridge with a truck” is a Boston slogan for the ages

@NeightFSU @oliver_drk @nypost No, the Chinese could not have engineered the virus to protect their own. (Not to mention that their own weren't in fact protected.) In terms of human genetics, this is complete nonsense

@oliver_drk @nypost Horribly worded. Could the Chinese had engineered it to protect their own? Absolutely. I’m sure other countries are doing the same. Could it be a coincidence that it also protected another group in the process? Yes. Doesn’t mean the Jews & Chinese are working together on that.

RFK Jr. says COVID was ‘ethnically targeted’ to spare Jews trib.al/1DuIwdD

@CentristAllWay @AlsoWonderWoman @lucky_number9_ @oliver_drk @nypost The virus has changed its binding to ACE2 repeatedly - it's the receptor binding domain that keeps changing between variants. There turn out to be many human variants that carry risk of disease, all populations have some, and the Chinese have shown no special immunity.

@AlsoWonderWoman @lucky_number9_ @oliver_drk @nypost The study talks about ACE2 receptors. The virus changed in it virulence and morbidity, not its binding to ACE2 receptors. You're absolutely right. The scientists at Cleveland Clinic and Case Western who wrote the paper are Anti-Semitic. We should try to cancel them.

@loco_qf @WiringTheBrain I had one last night. I was using my phone to take video of the weird array of animals that had wandered into the room when I remembered that there is some kind of blob on the lens (which there is)

Found out last night that I can taste and smell food in my dreams (made a delicious dream risotto). I'm wondering how common this is:

@sluntz @PGeldsetzer1 @guardian Over 80, not over 70

@PGeldsetzer1 @guardian Fascinating research. I'm also bewildered how they justified preventing people over 70 from having access to the vaccine - seems incredibly discriminatory.

Biggest thing to ever come out of my little group. Pls help spread this finding! We found clean, CAUSAL evidence that the shingles vaccine prevents a good chunk of dementia cases. So, could a virus cause Alzheimer’s->YES! Hear me out & see preprint: bit.ly/3MVqXU9 🧵1/

@ProfRobHoward @erikacule @Foreman1David @Rational_Psych @PGeldsetzer1 @guardian And yet you're willing to criticize the study publicly without having read beyond the abstract or understood the study design

@erikacule @Foreman1David @Rational_Psych @PGeldsetzer1 @guardian If I had what I thought was an exciting finding, I’d wait till it had gone through appropriate peer review and was published in a good journal before speaking to journalists about it.

@JonathanMTweet @aylwyn_scally The common ancestor of mammals split into two branches. The issue is, which later animals come from those two branches. According to this study, one branch leads to comb jellies and the other to all other animals.

@aylwyn_scally Would it be OK to rephrase it as something like, "which living species is genetically closest to the the very first animals"?

Unless the proposal is that animals evolved more than once (and it isn't), this is incorrect. Every animal today is an equally close relative of the first animals.

@JonathanMTweet @aylwyn_scally No. All living animals have been evolving for exactly the same length of time since the 1st animal. Some lineages do accumulate mutations faster, but that's not what's meant here. (By that measure, humans are closer to that first animal than mice, say.)

@smerdon_bonnie @yougotvitchard @ICeeCoo @ClaraJeffery @propublica No, you have to review the judicial rules to determine whether the judicial rules have been violated. You just have to be paying attention to know that something wrong occurred.

@drugmonkeyblog There's a distribution. Some I've been an author on are probably in the 1000+ bucket while for others 10-20 is wildly optimistic.

When you publish a scientific paper approximately how many people do you expect to read it carefully and in depth?

@baym I had one but it was only visible to those who were worthy.

Oh yeah? Well my blue check was never there to begin with