Martin Brümmer

Wintergartenhochhaus, Leipzig, Germany The Wintergartenhochhaus is a 95 meter tall skyscraper in Leipzig, Germany. It is currently the cities third tallest building. When construction was finished in 1972 it was the tallest residential bui...

Credit for the model goes to Tralux21: rebrickable.com/mocs/MOC-238...

WIBET (Wouldn’t It BE Tight) science

The recording is now available so that you can confirm that I indeed have a German accent and color-match my outfits with my Zoom background.

youtu.be/YL0co26ng-g?...

Can confirm, I’m happy to support your powerful message with top notch production value ❤️

Thank you so much! It’s a Sony Alpha 6000 on an Elgato Camlink 4K, an Elgato Keylight, a Rohde ProCaster on a Focusrite Scarlett Solo. In the background a Twinkly Squares panel and a Milky Way Lego, lit by a Philips Hue gradient light strip. Front and center, @dingdingpeng.the100.ci (priceless).

Well no. Ring bearers and elves went to Valinor because the age of men had come, right?

So it's actually postdocs who will reign over middle academia until the Gift of Ilúvatar frees then from the confines of Science.

When we first met, I fell in love with your beautiful eyes... all 50 of them!

The threat of analytic flexibility in using large language models to simulate human data: A call to attention Social scientists are now using large language models to create "silicon samples" - synthetic datasets intended to stand in for human respondents, aimed at revolutionising human subjects research. How...

Can large language models stand in for human participants?
Many social scientists seem to think so, and are already using "silicon samples" in research.

One problem: depending on the analytic decisions made, you can basically get these samples to show any effect you want.

THREAD 🧵

Meine Erwartung wäre, dass hierdurch die Nachfrage steigt, die Anbieter daraufhin das Angebot steigern, ihre Produktion also skalieren und so der Preis langfristig fällt.

This has now been published in PNAS!

www.pnas.org/doi/10.1073/...

The original authors also posted a reply:

www.pnas.org/doi/10.1073/...

Quick thread on some additional thoughts and then I'm probably done talking about this one 🧵

Turned out that he was right insofar simply not using GitHub Actions easily fixed the problem.

We are mainly an enterprise software. The open source community is largely locked into GitHub because of the network effect, not because it is inherently a better solution.

Just finished "The Collapse" about the precise events that led up to the opening of the Berlin Wall -- struggle within the Soviet Bloc and demonstrations in Saxony (most prominently Leipzig), bureaucratic miscommunication, and the chaotic and improbable events in the night itself.>

You jest, but while we can discuss if specific products are good or bad products, asserting there’s ill intentions behind most of the digital technology we are using is just wrong. They are made to solve problems and make money and there’s nothing wrong with that.

Please describe to me how GitLab is evil. Or Kubernetes. Or tensorflow. Or RStudio. Or Google docs even. It’s just a bad statement in its scope and you know it.

This is incredibly reductive and paints all digital infrastructure and the associated productivity gains as deliberately morally reprehensible. You are saying that hundreds of thousands of engineers that work to keep the tools you need to do your job online are evil people.

Patient-Specific In Vivo Gene Editing to Treat a Rare Genetic Disease | NEJM Base editors can correct disease-causing genetic variants. After a neonate had received a diagnosis of severe carbamoyl-phosphate synthetase 1 deficiency, a disease with an estimated 50% mortality ...

1. Researchers at the Children’s Hospital of Philadelphia diagnosed a baby with a life-threatening genetic disease within days of birth, and spent 6 months designing & testing a personalized gene therapy to fix his deficiency.

He could reduce medications, gain weight and move up on the growth chart

I love how I make a random joke one evening and find my vision realized in shitpost form a day later. Very convenient, thank you for your service!

I would love to tell you that it’s easier in GitLab but that would be a lie.

She said while looking for (and finding!) the right stain remover for her stain. Who’s laughing now?

Just measure their weight over time to plot the rate of decay

Why it's okay to use ChatGPT, in 4 graphs

It took more than a decade of arguing over p-values and replicability -- but it's inspiring to see that meta-science in psychology is finally tackling important questions.

a man with a mustache and a blue jacket is holding a glass of wine . Alt: a man with a mustache and a blue jacket is throwing confetti

Thrilled to announce that I’m starting 2025 with a new editorial role—excited to serve as a Senior Editor at Psychological Science!

The Untold Mystery of Rogue RA The 100% CI unravels serial cases of mysterious research misconduct that can be traced back to a single villain striking from the shadows: Rogue RA

New post on The 100% CI! We unravel a series of mysterious research misconduct cases that can be traced back to a single villain striking from the shadow: Rogue RA

www.the100.ci/2024/12/18/r...

We can publish a Gitbook using GitLab pages. Just a bunch of markdown files in a public repository, easy to edit and comment on.

Pandemics have killed millions of people throughout history.

How many deaths were caused by major pandemics, and how have researchers estimated their death tolls?

New article by me!
ourworldindata.org/historical-p...

Thank you for asking and feel free to share! Note that it’s full of assets that are copyrighted by someone else, although I would argue fair use here.