Lightnews — Scholar-powered news

Reposted by Darren Dahly

Sarah Karlin-Smith @sarahkarlin-smith.bsky.social · 2d

Vinay Prasad has a new deputy from Arnold Ventures world and George Tidmarsh is comparing FDA reform to Dresden...

What could it mean...

insights.citeline.com/pink-sheet/a...

FDA Leadership Philosophies: CBER Promotions Spark Concerns; Tidmarsh Says He’s Not A Destroyer

New promotions and social media posts may offer some hints as to the philosophies guiding agency leaders. One CBER appointee stands out for her connections to Arnold Ventures.

insights.citeline.com

1 3 5

Reposted by Darren Dahly

Michael Hobbes @michaelhobbes.bsky.social · 1d

There must be some mistake, I read 50 Atlantic articles about how RFK Jr just wants to help Americans eat more vegetables

Gravel Influencer @gravelinfluencer.bsky.social · 1d

They continue to eviscerate the public health machinery that has quietly kept Americans, and others around the world, safe and healthy for decades.

The staff of the Morbidity and Mortality Weekly Report, the journal that reports on health trends and emerging infectious threats, was also laid off. The publication’s storied history includes a June 1981 report that five previously healthy gay men were treated for an unusual pneumonia — the first hint of the AIDS epidemic.

Roughly 70 Epidemic Intelligence Service officers — the so-called “disease detectives” who respond to outbreaks around the globe — received layoff notices, according to a person familiar with them. The service was spared during an earlier round of layoffs in February.

17 260 1.6K

Reposted by Darren Dahly

Hisham Zerriffi @hishamzerriffi.bsky.social · 11h

“Your job is not to turn in completed assignments; it's to learn how to think.”

Ted Chiang‘s succinct summary of the issue with using AI in education.

I’m not a sports guy but the athlete metaphor he uses in this interview is pretty spot on.

The Incompatibilities Between Generative AI and Art: Q&A with Ted Chiang

cdh.princeton.edu

1 23 54

Reposted by Darren Dahly

George Davey Smith @mendelrandom.bsky.social · Sep 12

"Triangulation: moving from correlations to causation", Cambridge School of Clinical Medicine Distinguished Lecture 2025, Thursday, October 16th, 5pm, in person or online
www.eventbrite.co.uk/e/school-of-...

13 19

Darren Dahly @statsepi.bsky.social · 19h

🙏👊

1

Darren Dahly @statsepi.bsky.social · 19h

What is lost if we stopped adding to the "training data" a thousand years ago, or a hundred, or ten, or even just last year? What new thing is missed in our urge to more efficiently produce endless old things?

28

Darren Dahly @statsepi.bsky.social · 19h

I read and write, I explore and I question, I design and script and analyse, I interpret and communicate. I do this to train my mind in the hopes of one day generating new knowledge. New knowledge that might even be useful, and that no algorithm can yet be trained on.

Terry McGlynn @hormiga.bsky.social · 4d

Y'all. I just got ChatGPT to do everything in R for this manuscript. I mean EVERYTHING. And it's all legit and reproducible. I'm shook.

How are we mentoring our trainees in statistics now? Who needs to learn coding in R line by line, and who doesn't?

scienceforeveryone.science/statistics-i...

Statistics in the era of AI

How do we mentor, teach, and do stats when AI can do so much of the work?

scienceforeveryone.science

4 13 62

Reposted by Darren Dahly

Ben Harrap @bharrap.bsky.social · 3d

I clearly had a lot of thoughts/feelings from reading Terry's blog post, so I wrote one of my own

It ended up being a bit of a stream-of-thought post but I think it captures some of my worries

benharrap.com/post/2025-10...

#statssky #episky #academicsky

5 28 77

Reposted by Darren Dahly

Dr David Robert Grimes @drg.bsky.social · 6d

Asked lots about this piece on Chronic Lyme disease, and disappointed in the @guardian (a publication I both subscribe and contribute to) for running it. Short thread explaining the multitude of problems with it, and why I consider it irresponsible to run as is..🧵

4 52 100

Reposted by Darren Dahly

Jonathan Howard @joho.bsky.social · 1d

Makary is now saying that Lyme came from a lab too.

youtu.be/bQovoQaaE4w?...

Silly Makary says COVID and Lyme came from labs.

YouTube video by Jonathan Howard

youtu.be

24 34 140

Reposted by Darren Dahly

spooky mental haunted house @jesseltaylor.bsky.social · 1d

Imagine if you did this in any other journalistic context

"Did the Tigers or Mariners Win Game 5? In Detroit, Depends on the Narrator."

David Corn @davidcorn.bsky.social · 1d

A classic case of irresponsible both-sidesism from the NYT. The story is that Trump and MAGA propagandists are lying about Portland to incite a conflict, not that there are different views of the matter.

14 160 530

Darren Dahly @statsepi.bsky.social · 1d

And good luck trying to change a med school curriculum or move away from mean/modes/medians/nhst/good luck! Statistics is special because more than any other topic, it must be "made easy" as I have been told more than once.

2

Darren Dahly @statsepi.bsky.social · 1d

Because it's not important. If it was important, educational institutions would ensure it was done well. Otherwise you are just rolling the dice that you are taught by some small % of academics in any field that are good at teaching AND supported to be good at teaching.

1 2

Darren Dahly @statsepi.bsky.social · 1d

To the degree that we want to diagnose and predict illness, or modify exposures to prevent/mitigate/reverse it, these concepts of a healthy microbiome or dysbiosis as described below have not, in my opinion, been useful.

2 2 10

Reposted by Darren Dahly

Angie Rasmussen @angierasmussen.bsky.social · 2d

I don’t know who needs to hear this but the CDC is being eviscerated right now. America is not going to have any kind of outbreak response capacity after tonight. Americans’ health data is no longer secure. Say goodbye to federal public health in any capacity. It’s a disaster. We won’t recover.

520 6.2K 14K

Darren Dahly @statsepi.bsky.social · 1d

Statisticians in my field have been trying to improve statistical literacy in earnest since the early 20th century. The @royalstatsoc.bsky.social has an entire section on teaching statistics. We have the tools. The knowledge simply isn't valued enough.

statsepi.substack.com/p/a-brief-hi...

A brief history of medical statistics and its impact on reproducibility.

You have to admit that medical research is a strange bird.

statsepi.substack.com

2 6 31

Darren Dahly @statsepi.bsky.social · 1d

Teaching scientists about statistics is not even that difficult. But it slows down how many papers get published, so we can't do it. I mean, why would we put any brakes on LLM produced papers by insisting scientists and reviewers understand the output!? Madness.

Dave Baltrus 🦦 @surtlab.bsky.social · 3d

As coding gets more plug and play, it becomes IMPERATIVE for us as educators to teach fundamentals about what the statistical tests are comparing and assumptions.

Students might not need to understand how to code things up, but they should understand what the snippets are doing and why

Terry McGlynn @hormiga.bsky.social · 4d

Y'all. I just got ChatGPT to do everything in R for this manuscript. I mean EVERYTHING. And it's all legit and reproducible. I'm shook.

How are we mentoring our trainees in statistics now? Who needs to learn coding in R line by line, and who doesn't?

scienceforeveryone.science/statistics-i...

6 9 47

Reposted by Darren Dahly

Wim Meijer @wimgmeijer.bsky.social · 1d

Nitazenes can be as much as 500 times more potent than opium-derived drugs.

This means that just a tiny amount can be deadly. There were 179 confirmed deaths from nitazene overdoses in the UK since May 2024.

Ireland too has seen several nitazene overdose deaths.

🧪

www.nature.com/articles/d41...

Dangerous ‘nitazene’ opioids are on the rise: researchers are worried

This class of synthetic opioids is more potent than heroin and morphine.

www.nature.com

6 5

Reposted by Darren Dahly

Emily Moin @emilymoin.com · 2d

I am open to the idea that there are people who don't have and don't want to gain the skills to engage directly with their data but every single day that I do I learn the answer to a question you'd never even think to ask unless you were personally staring into the abyss of an uncleaned dataset.

5 11 79

Reposted by Darren Dahly

Prof Christina Pagel @chrischirp.bsky.social · 3d

🧵🚨

The UK’s independent scientific bodies are highly vulnerable to politicisation - over the past 5 months I've been working with @martinmckee.bsky.social to map out their vulnerabilities and it's not good news.

Today our report is published!
www.ucl.ac.uk/policy-lab/n...

1/11

UK’s arm’s length public bodies are highly vulnerable to politicisation

Seven in ten Britons say it is important for top scientific institutions to be independent in exclusive new polling.

www.ucl.ac.uk

20 410 630

Reposted by Darren Dahly

Reese Richardson @reeserichardson.bsky.social · 2d

Hope to see you there! ICYMI: thousands of design registrations have been sold as “patents” as an explicit ploy for academics to pad their CVs. This is still happening! For example, here's a design registration filed 20 Sep 2025 for "MACHINE LEARNING BASED SECURITY DEVICE FOR CLOUD COMPUTING":

A green computer monitor mounted to a box. The box features numerous buttons and panels, oversized ethernet ports and a panel of every type of USB recievers. The whole assembly is mounted on four spindly-looking robot legs with wheels.

1 2 9

Reposted by Darren Dahly

Crystal Lewis @cghlewis.bsky.social · 3d

I am so grateful for everyone who has subscribed (500 of you), liked, shared, promoted, and bought me coffee in support of RDM Weekly! Just knowing that people are reading is a huge win. I have really enjoyed putting this newsletter together over the last 3 months and look forward to continuing it!

2 3 20

Reposted by Darren Dahly

Charles Driver @charlesdriver.bsky.social · 2d

Some more shiny apps for the classroom -- here looking at different approaches to pooling information across groups (e.g., partial pooling), and how the group level predictions shift with changes in amount of data used for fitting.

charlesdriver.shinyapps.io/Pooling_Info...

Exploring Pooling of Information Across Groups

charlesdriver.shinyapps.io

1 3 13

Reposted by Darren Dahly

Prof Gavin Yamey @gavinyamey.bsky.social · 2d

🚨In a new piece that I co-authored with @michaeldgreen.com for @bmj.com, we argue that Trump's war on DEI in medical education will damage the health of Americans

We present research evidence showing that a diverse medical workforce improves population health outcomes

www.bmj.com/content/391/...

Trump’s war on diversity in medical education could shorten Americans’ lives

Building a diverse medical workforce is a crucial step to improving population health outcomes, write Gavin Yamey and Michael D Green The Trump administration and its allies have recently taken seve...

www.bmj.com

11 25

Reposted by Darren Dahly

Hannah Daly @hannahdaly.ie · 2d

The full quote as reported:

“I don’t think we can mitigate for climate change,” he said.
“This might be controversial, because I believe in addressing climate change....

Joe Noonan @joenoonan.bsky.social · 3d

“I don’t think we can mitigate for climate change.”

- Mícheál Martin, Taoiseach.

www.independent.ie/business/tec...

Taoiseach says some climate projects will not be fulfilled because they risk polarising society

Taoiseach Micheál Martin has rowed back on climate change commitments, saying Ireland may not be able fulfil all current initiatives due to the polarising effect it could have on society.

www.independent.ie

2 20 40